About our tests

Overview

IIHS evaluates a vehicle's crashworthiness with the help of six tests: moderate overlap front, driver-side small overlap front, passenger-side small overlap front, side, roof strength and head restraints & seats. For front crash prevention ratings, the Institute conducts low- and moderate-speed track tests of vehicles with automatic braking systems. IIHS also conducts evaluations of headlight systems and of the child seat attachment hardware known as LATCH. The descriptions below explain how each test is conducted and how the results translate into ratings.

For the crash tests we conduct, we purchase the vehicles from dealers just like an ordinary consumer. If the test is being conducted at the manufacturer's request and isn't part of our regular testing schedule, then the manufacturer reimburses us for the vehicle.

Transparency is one of the keys to the success of our ratings program. All of our test protocols are available here.

Frontal crash tests

A frontal crash is the most common type of crash resulting in fatalities. Major strides have been made in frontal protection, thanks in large part to the crash test program that the National Highway Traffic Safety Administration (NHTSA) began in the late 1970s and the crashworthiness evaluations that IIHS began in 1995.

IIHS conducts three different frontal crash tests: a moderate overlap test (formerly known as the frontal offset test), a driver-side small overlap test and a passenger-side small overlap test.

Moderate overlap test overhead shot
Moderate overlap frontal test configuration
Driver-side small overlap test overhead shot
Driver-side small overlap frontal test configuration
Passenger-side small overlap test overhead shot
Passenger-side small overlap frontal test configuration

Moderate overlap frontal test

When we began our moderate overlap frontal tests, the majority of vehicles were rated poor or marginal. Today, all vehicles earn good ratings. Occupant compartments are much stronger than they used to be. They hold up in a crash and allow safety belts and airbags to do their jobs.

In the moderate overlap frontal test, a vehicle travels at 40 mph toward a barrier with a deformable face made of aluminum honeycomb. The barrier face is just over 2 feet tall. A Hybrid III dummy representing an average-size man is positioned in the driver seat. Forty percent of the total width of the vehicle strikes the barrier on the driver side.

The forces in the test are similar to those that would result from a frontal offset crash between two vehicles of the same weight, each going just under 40 mph.

Driver-side small overlap frontal test

To help encourage further improvements in frontal crash protection, the Institute in 2012 introduced a driver-side small overlap frontal crash test. The test is designed to replicate what happens when the front left corner of a vehicle collides with another vehicle or an object like a tree or utility pole. This crash test is a challenge for some safety belt and airbag designs because occupants move both forward and toward the side of the vehicle.

In the driver-side small overlap frontal test, a vehicle travels at 40 mph toward a 5-foot-tall rigid barrier. A Hybrid III dummy representing an average-size man is positioned in the driver seat. Twenty-five percent of the total width of the vehicle strikes the barrier on the driver side.

Most modern cars have safety cages encapsulating the occupant compartment and built to withstand head-on collisions and moderate overlap frontal crashes with little deformation. At the same time, crush zones help manage crash energy to reduce forces on the occupant compartment. The main crush-zone structures are concentrated in the middle 50 percent of the front end. When a crash involves these structures, the occupant compartment is protected from intrusion, and front airbags and safety belts can effectively restrain and protect occupants.

Small overlap frontal crashes primarily affect a vehicle's outer edges, which aren't well protected by the crush-zone structures. Crash forces go directly into the front wheel, suspension system and firewall. It is not uncommon for the wheel to be forced rearward into the footwell, contributing to even more intrusion in the occupant compartment and resulting in serious leg and foot injuries. To provide effective protection in small overlap crashes, the safety cage needs to resist crash forces that aren't tempered by crush-zone structures. Widening these front-end structures also helps.

Passenger-side small overlap frontal test

Manufacturers have responded to the driver-side small overlap test by improving vehicle structures and airbags, and most vehicles now earn good ratings. However, IIHS research tests demonstrated that those improvements didn't always carry over to the passenger side. Discrepancies between the left and right sides of vehicles spurred us to develop a passenger-side small overlap test and begin issuing passenger-side ratings in 2017.

The passenger-side test is the same as the driver-side test except the vehicle overlaps the barrier on the right side. In addition, instead of just one Hybrid III dummy, there are two — one in the driver seat and one in the passenger seat.

Ratings criteria

Engineers consider three factors to determine how a vehicle rates in the moderate overlap and small overlap frontal tests: structural performance, injury measures and dummy movement.

Structure/safety cage: To assess a vehicle's structural performance, engineers measure the amount of intrusion into the occupant compartment at key locations in the interior and exterior of the vehicle after the crash. The amount and pattern of intrusion shows how well the front-end crush zone managed the crash energy and how well the safety cage held up.

Injury measures: Sensors in the dummy — or dummies, in the case of passenger-side small overlap — are used to determine the likelihood that a driver or passenger would sustain various types of injuries in a similar real-world crash. Measures recorded by sensors in the head, neck, chest, legs and feet of the dummy indicate the level of stress or strain on that part of the body — in other words, the risk of injury.

Restraints/dummy movement: Even if injury measures are low, it's important to consider the movement of the dummy or dummies during the crash, since not all occupants are the same size as the dummies or seated exactly the same way. A close call for a dummy could be an actual injury for a person.

Before each crash test, technicians put greasepaint on the dummy's head, knees and lower legs. After the test, the paint shows what parts of the vehicle came into contact with those parts of the dummy. The paint, combined with high-speed film footage of the crash, allows engineers to evaluate the movement of the dummy or dummies.

Understanding the ratings

How do vehicles that earn good ratings in the moderate overlap frontal test perform in similar real-world crashes? An analysis of 14 years worth of crash data involving IIHS-rated vehicles shows that a driver of a vehicle rated good in the moderate overlap test is 46 percent less likely to die in a frontal crash, compared with a driver of a vehicle rated poor. A driver of a vehicle rated acceptable or marginal is 33 percent less likely to die than a driver of a poorly rated one. Since the Institute only started evaluating vehicles for protection in small overlap frontal crashes in 2012, we don't have comparable data on how vehicles with good small overlap ratings fare in the real world.

Frontal crash test results can't be used to compare vehicle performance across weight classes. That's because the kinetic energy involved in the moderate overlap and small overlap frontal tests depends on the speed and weight of the test vehicle. Thus, the crash is more severe for heavier vehicles.

Given equivalent frontal ratings, the heavier of two vehicles usually offers better protection in real-world crashes. In 2009, IIHS demonstrated this principle with a series of tests in which small cars were crashed into larger cars, all of which had good frontal ratings in the moderate overlap test.

For information about how ratings are kept up-to-date from one model year to the next, see our test verification information.

How do the Institute's frontal crash tests differ from NHTSA's New Car Assessment Program frontal test?

In the New Car Assessment Program (NCAP), passenger vehicles are crashed at 35 mph into a rigid barrier that covers the full width of the vehicle.

The Institute runs offset frontal tests instead of full-width frontal tests. In an offset crash only one side of a vehicle's front end, not the full width, hits the barrier. As a result, a smaller part of the structure has to manage the crash energy, and intrusion into the occupant compartment is more likely. An offset test is more demanding of a vehicle's structure than a full-width test, while a full-width test is more demanding of safety belts and airbags. In a full-width test, there is less crushing of the vehicle structure so the decelerations that these restraints must handle are greater. Together, the tests provide a more complete picture of frontal crashworthiness than either test by itself.

NHTSA full-width test image
NHTSA's full-width frontal test configuration

NCAP has been extremely successful. Automakers responded by improving their vehicles to earn good ratings, giving them a higher level of frontal crashworthiness than is required by law. By the 1990s, most vehicles were performing well in the full-width frontal test.

In 2010, NHTSA announced an enhanced 5-star rating system that toughened criteria and combined ratings from its different tests into a single rating to provide consumers an overall view of vehicle safety. The NCAP changes apply to 2011 and later models.

Side crash test

Side crashes account for about a quarter of passenger vehicle occupant deaths in the United States. Protecting people in side crashes is challenging because the sides of vehicles have relatively little space to absorb energy and shield occupants, unlike the fronts and rears, which have substantial crumple zones. Automakers have made big strides in side protection by installing side airbags and strengthening the structures of vehicles. The Institute's testing program has played a key role in bringing about these improvements.

Side airbags, which today are standard on most new passenger vehicles, are designed to keep people from colliding with the inside of the vehicle and with objects outside the vehicle in a side crash. They also help by spreading impact forces over a larger area of an occupant's body. However, side airbags by themselves are not enough. Strong structures that work well with the airbags also are crucial.

Need for side testing

IIHS began its side test program in 2003. At that point, the federal government was already performing side tests on new passenger vehicles as part of the New Car Assessment Program. But we were concerned that the government's test didn't completely capture the types of crashes likely to occur in the real world.

IIHS vs. NHTSA barrier height
NHTSA barrier shown in yellow, superimposed
over the taller IIHS barrier

That's because the moving barrier used in the government's test was developed in the early 1980s, when most of the vehicles on the road were cars, before SUVs and pickups became as prevalent as they are today. The height of the barrier's front end is below the heads of the crash test dummies. As a result, the federal test doesn't assess the much greater risk of head injury from impacts with taller vehicles. To fill this gap, we initiated our own test with a different barrier — one with the height and shape of the front end of a typical SUV or pickup.

How the test works

In the Institute's test, a 3,300-pound SUV-like barrier hits the driver side of the vehicle at 31 mph. Two SID-IIs dummies representing small (5th percentile) women or 12-year-old children are positioned in the driver seat and the rear seat behind the driver.

IIHS was the first in the United States to use this smaller dummy in a test for consumer information. It was chosen because women are more likely than men to suffer serious head injuries in real-world side impacts. Shorter drivers have a greater chance of having their heads come into contact with the front end of the striking vehicle in a left-side crash.

Our side test is severe. It's unlikely that people in comparable real-world crashes would emerge uninjured. With good side protection, however, people should be able to survive a crash of this severity without serious injuries.

Ratings criteria

Engineers look at three factors to determine side ratings: driver and passenger injury measures, head protection and structural performance.

Injury measures: Injury measures from the two dummies are used to determine the likelihood that occupants would sustain significant injuries in a real-world crash. Measures are recorded from the head, neck, chest, abdomen, pelvis and femur. These injury measures, especially the ones from the head and upper body, are major components of each vehicle's overall side rating.

Dummy head with greasepaint
A technician applies greasepaint before a crash test.
Smeared greasepaint on airbag
Smeared greasepaint shows where the driver dummy's head hit the side curtain airbag.

Head protection: To supplement head injury measures, technicians put greasepaint on the dummies' heads before each crash test. After the test, the paint shows what parts of the vehicle or the barrier came into contact with the heads. If the vehicle has airbags and they perform correctly, the paint should end up on them.

In cases when the barrier hits a dummy's head during impact, the dummy usually records very high injury measures. That might not be true, however, with a "near miss" or a grazing contact. The paint, along with footage of the test recorded on high-speed film, helps identify such cases, which is important because small differences in occupants' heights or seating positions compared with those of the test dummies could result in a hard contact and high risk of serious head injury.

Structure/safety cage: Engineers assess the vehicle's structural performance by measuring the amount of intrusion into the occupant compartment around the B-pillar (between the doors). Some intrusion into the occupant compartment is inevitable in serious side impacts, but it shouldn't seriously compromise the driver and passenger space. As with head protection, this is another assessment that helps evaluate the injury risk of occupants who aren't exactly the same size or sitting in exactly the same positions as the dummies.

Understanding the ratings: How much better are vehicles that earn good ratings?

In the real world, a driver of a vehicle rated good is 70 percent less likely to die in a left-side crash, compared with a driver of a vehicle rated poor. A driver of a vehicle rated acceptable is 64 percent less likely to die, and a driver of a vehicle rated marginal is 49 percent less likely to die.

Those numbers come from an analysis of a decade's worth of crash data on Institute-rated vehicles. Only vehicles with standard side airbags were included, and the results demonstrate that just having airbags doesn't guarantee good protection. Our tests show how airbags and a vehicle's structure work together in an actual crash. If the occupant space remains largely intact, then the safety belts and side airbags have time to control the motion of the crash test dummies and keep injury measures low. That's less likely to happen if the side of the vehicle is significantly crushed.

Unlike frontal crash test ratings, side ratings can be compared across vehicle type and weight categories. This is because the kinetic energy involved in the side test depends on the weight and speed of the moving barrier, which are the same in every test. In contrast, the kinetic energy involved in the frontal crash test depends on the speed and weight of the test vehicle.

When side airbags are optional, we test the vehicle without the option. A second test is conducted with the optional airbags if the manufacturer requests it and reimburses the Institute for the cost of the vehicle. Both results are published on the website. In recent years it has become rare for any vehicle not to have standard airbags.

For information about how ratings are kept up-to-date from one model year to the next, see our test verification information.

Tests drive progress

2008 vs. 2005 Lancers

GOOD: 2008 Mitsubishi Lancer with side airbags (top)

POOR: 2005 Mitsubishi Lancer without side airbags (bottom)

When we began side testing in 2003, only about 1 of 5 vehicles tested earned good ratings. Nearly all the others were rated poor.

Since then, airbags have become standard equipment in the vast majority of passenger vehicles, and occupant compartments have become much stronger. These changes are in large part a direct result of the Institute's testing program. Manufacturers know consumers consult the ratings before buying, so they design vehicles with our tests in mind. As a result, most current vehicle designs earn good ratings.

How does the Institute's side crash test differ from NHTSA's side tests?

Since our side crash program was launched, NHTSA has made changes to its New Car Assessment Program. It continues to conduct its original side crash test representing a T-bone crash and using a moving barrier. In 2010, NHTSA added a side pole test to address side crashes in which vehicles slide into fixed objects. Unlike NHTSA's moving-barrier test, the pole test requires protection for the driver's head. Both NHTSA tests evaluate vehicle performance solely on measures from the dummies' sensors. 

The IIHS side test typically causes more damage than either of the NHTSA tests. We directly evaluate structural deformation in addition to the measures from dummy sensors. We also review test video to check  that structures and airbags work well together to protect occupants.

In the NHTSA moving barrier crash test, the barrier, which resembles the front of a car from the 1970s, strikes the side of the test vehicle at 39 mph and an angle of 27 degrees. Although the test speed is higher than the IIHS test, the lower barrier shape pushes against the strong lower portion of the door frame that is often missed in real crashes. This makes it easier for a vehicle's structure to protect occupants than in the Institute’s test, in which the SUV-like barrier misses that structure except in the case of taller vehicles.

In NHTSA's pole test, the test vehicle slides sideways at 20 mph and a 75-degree angle into a 10-inch-diameter rigid pole. A SID-IIs dummy is in the driver seat, but rear seat head protection is not tested. In contrast, our single, moving-barrier test encourages good head protection for both front- and rear-seat occupants.

Roof strength test

Thousands of people are killed each year in rollovers. The best way to prevent these deaths is to keep vehicles from rolling over in the first place. Electronic stability control is significantly reducing rollovers, especially fatal single-vehicle ones. When vehicles do roll, side curtain airbags help protect the people inside, and belt use is essential. However, for these safety technologies to be most effective, the roof must be able to maintain the occupant survival space when it hits the ground during a rollover. Stronger roofs crush less, reducing the risk that people will be injured by contact with the roof itself. Stronger roofs also can prevent occupants, especially those who aren't using safety belts, from being ejected through windows, windshields or doors that have broken or opened because the roof has deformed.

In the test, the strength of the roof is determined by pushing an angled metal plate down on one side of the roof at a slow but constant speed and measuring the force required to crush the roof. The force applied relative to the vehicle's weight is known as the strength-to-weight ratio. The peak strength-to-weight ratio recorded at any time before the roof is crushed 5 inches is the key measurement of roof strength.

A good rating requires a strength-to-weight ratio of at least 4. In other words, the roof must withstand a force of at least 4 times the vehicle's weight before the plate crushes the roof by 5 inches. For an acceptable rating, the minimum required strength-to-weight ratio is 3.25. For a marginal rating, it is 2.5. Anything lower than that is poor.

The figure below shows sample results for two vehicles — one rated good and one rated poor. Peak force for Vehicle A is 7.26. Since that number is higher than 4, the vehicle is rated good. Peak force for Vehicle B is 2.31. Since that number is lower than 2.5, the vehicle is rated poor.

Strength-to-weight-ratio graph image

The following video shows how the roof strength test is conducted. In this test of the 2010 Buick LaCrosse, the peak force is 19,571 pounds for a strength-to-weight ratio of 4.90 and a good rating. The playback speed of this video has been increased. The plate normally crushes at a rate of about 1/8 inch per second.

In every test, the roof is crushed 5 inches. What varies — and can't be seen in a video — is the force used by the machine to achieve that degree of crush. To demonstrate how roof strength can vary and what those differences mean for people inside a vehicle during a rollover, IIHS conducted a demonstration in which two vehicles with different roof strength ratings were subjected to identical force. This video shows what happened when the 2009 Volkswagen Tiguan, rated good for roof strength, and the 2008 Kia Sportage, rated poor, were each subjected to a crush force of 15,000 pounds.

Head restraints & seats test

Neck sprains and strains are the most frequently reported injuries in U.S. auto insurance claims. Such whiplash injuries can be sustained in any type of crash but occur most often in rear-end collisions. Good head restraints can help prevent them.

IIHS tests vehicle seats and head restraints with a special dummy that has a realistic spine. The vehicle seat — with the dummy in it — is placed on a sled, which is moved to simulate a rear impact.

Head restraint geometry

Good geometry is essential for an effective head restraint. If a head restraint isn't behind and close to the back of an occupant's head, it can't prevent whiplash in a rear-end collision. IIHS evaluates the geometry of head restraints in passenger vehicles based on the height and backset relative to an average-size male. A restraint should be at least as high as the head's center of gravity, or about 3.5 inches below the top of the head. The backset, or distance behind the head, should be as small as possible. Backsets of more than about 4 inches have been associated with increased symptoms of neck injury in crashes. The restraints are measured with the angle of the torso at about 25 degrees, a typical seatback angle. IIHS classifies each restraint according to its height and backset into one of four geometric zones — good, acceptable, marginal or poor.

Head restraints have improved since the Institute first began rating them. In 1995, only 3 percent of measured head restraints received good geometric ratings, while 82 percent were rated poor. Among recent models, most head restraints have good geometric ratings, and the rest are acceptable. IIHS ratings have forced manufacturers to pay attention to head restraint design. More recently, a government standard in effect since September 2010 requires a minimum of 29.5 inches from an occupant's hip to the top of a head restraint and a backset of 2.2. inches or less. This guarantees that nearly all new head restraints would be rated good for geometry by IIHS.

Dynamic ratings derived from seat parameter and neck force results

Seat parameters + Neck forces = Dynamic ratings
Pass + Low =
G
Moderate
A
High
M
Fail + Low =
A
Moderate
M
High
P

Overall ratings derived from both geometric and dynamic ratings

Geometric rating + Dynamic rating = Overall rating
G
+
G
=
G
A
A
M
M
P
P
A
+
G
=
A
A
A
M
M
P
P
M
+ No dynamic test =
P
P
+ No dynamic test =
P

Dynamic ratings

Seats and head restraints with geometry rated good or acceptable are tested in a simulated rear impact conducted on a sled. The test assesses how well the seats support the torso, neck and head of a BioRID dummy. The test simulates a rear-end crash with a velocity change of 10 mph, approximately equivalent to a stationary vehicle being struck at 20 mph by a vehicle of the same weight.

A seat/head restraint's dynamic rating depends on performance in the sled test. There are two sets of criteria for evaluating performance. The first criteria are the two seat design parameters, time to head restraint contact (must be ≤70 milliseconds to pass) and torso acceleration (must be ≤9.5 g to pass). The second set of evaluation criteria is comprised of the maximum neck shear force and maximum neck tension measured on the dummy during the test. These neck forces (classified as low, moderate or high) indicate how well or how poorly an occupant's head and neck would be supported in a rear impact at low to moderate speed. A seat that passes at least one of the seat design parameters and has low neck forces earns a dynamic rating of good.

Overall ratings

The geometric rating and the dynamic rating are combined to produce a seat/head restraint combination's overall evaluation. A good rating can only be earned with both a good geometric rating and a good dynamic rating. Seats with only acceptable geometry earn an acceptable overall rating even if their ratings in the dynamic test are good. Seats rated marginal or poor for geometry would not be tested dynamically and automatically would be assigned a rating of poor; however, since the 2011 model year, all seats have had good or acceptable geometry.

The IIHS rating applies only to the specific seat/head restraint combination tested, though many vehicles have multiple options for these components. For a given vehicle, the Institute typically tests the seat option most likely to be found on dealer lots.

BioRID

BioRID 50th percentile male rear-impact dummy

Dummy and sled used in dynamic tests

Dynamic testing of seats/head restraints requires a dummy with a realistic spine and neck. Before the development of BioRID, or biofidelic rear impact dummy, all dummies had rigid spines and necks that didn't interact with vehicle seats the way human spines and necks do. BioRID was developed for rear testing by a consortium of Chalmers University, restraint maker Autoliv, Saab and Volvo. This dummy, representing an average-size man, has a spine composed of 24 articulated vertebra-like pieces. The spine interacts with vehicle seats during tests in much the same way as a human spine would. Plus BioRID's segmented neck can produce the motion observed by human necks in real-world crashes in which vehicles are struck from behind.

The device on which dynamic tests of seats/head restraints are conducted is a steel flatbed sled that runs on fixed rails. The sled is moved to simulate vehicle crash accelerations, re-creating the forces on occupants inside vehicles during real-world crashes. The changing acceleration or deceleration over the time duration of a crash is referred to as a crash pulse, and the key aspect of a sled is that it can be programmed to produce specific crash pulses. To evaluate rear crash protection, vehicle seats are affixed to the sled, which is accelerated to simulate a stationary vehicle that's rear-ended by another vehicle of the same weight going 20 mph. To accomplish this, compressed air is pumped into a special cylinder, thrusting a ram forward in a preprogrammed pattern of acceleration (crash pulse). Peak acceleration in the sled test is 10 g (5 g mean acceleration), and the duration is 91 milliseconds.

Understanding the ratings: How much better are vehicles with good head restraints?

Vehicles with good rear crash ratings do a better job of preventing neck injuries in the real world. In 1999, IIHS researchers analyzed more than 5,000 insurance claims and determined that drivers with head restraints with good geometric ratings were 24 percent less likely than drivers with poor-rated head restraints to sustain neck injuries in rear-end crashes.

A 2008 study looked at real-world crashes involving seat/head restraint combinations that had been dynamically tested. Insurance claims for drivers of cars and SUVs struck from behind were examined for evidence of driver neck injury. Injury rates were 15 percent lower for vehicles with seats/head restraints rated good compared with vehicles with seats/head restraints rated poor. Long-term injuries, or those lasting three months or more, were 35 percent lower for vehicles with seats/restraints rated good compared with seats/restraints rated poor.

Front crash prevention tests

IIHS launched the front crash prevention rating program in 2013 to help consumers zero in on the most effective systems. We began with ratings of systems that prevent vehicle-to-vehicle, front-to-rear crashes after research by HLDI indicated that forward collision warning and autobrake systems help drivers avoid front-to-rear crashes at both low speeds and moderate speeds.

Subsequent HLDI research revealed a benefit for front crash prevention systems that recognize pedestrians, and in 2019 a separate evaluation was added for systems that prevent vehicle-to-pedestrian crashes.

Models with optional or standard front crash prevention systems are rated as superior, advanced or basic in both the vehicle-to-vehicle evaluation and the vehicle-to-pedestrian evaluation.

Vehicle-to-vehicle ratings are determined by how the system performs in tests at 12 and 25 mph. The availability of forward collision warning also is factored in.

For a superior rating, a vehicle must have an autobrake system that can avoid a crash or substantially reduce speeds in both tests. For an advanced rating, a vehicle must have autobrake and avoid a crash or reduce speeds by at least 5 mph in 1 of 2 tests. Vehicles that have a warning system only earn a basic rating, provided the system meets National Highway Traffic Safety Administration performance criteria.

For vehicle-to-pedestrian ratings, scores in six tests — three scenarios run at two different speeds each — are combined to assign a rating of superior, advanced or basic.

How the vehicle-to-vehicle tests work

IIHS evaluates the stopping capabilities of vehicles equipped with autobrake in two tests at 12 and 25 mph on the Vehicle Research Center test track. In each, an engineer drives the vehicle straight toward a stationary target designed to simulate the back of a car.

Since running into an actual car puts the test driver at risk and is expensive, IIHS uses an inflatable target as a stand-in. Under the vinyl cover, inflatable tubes and foam sit on a metal frame, which is then affixed to metal guides on the track to keep the target from moving until it is struck by the test vehicle. A GPS system and other sensors monitor the test vehicle's lane position, speed, time to collision, braking and other data. An onboard camera captures each test run from the driver's perspective and monitors any warnings issued by the front crash prevention systems.

The Institute awards points based on how much the systems slow the vehicle to avoid hitting the target or lessen the severity of the impact in the two tests. In the case of an unavoidable collision, lowering the striking vehicle's speed reduces the crash energy that vehicle structures and restraint systems have to manage. That reduces the amount of damage to both vehicles involved in the collision and minimizes injuries to people traveling in them.

To earn a point for forward collision warning, the system must meet NHTSA criteria. That means the system must issue a warning before a specified time in 5 of 7 test trials under three scenarios. The agency identifies vehicles with systems that meet the standard as part of its online ratings.

How the vehicle-to-pedestrian tests work

Pedestrian crash prevention capabilities are evaluated using dummies that move across or stand in the roadway. Tests are conducted in these three scenarios:

Pedestrian test scenarios

Perpendicular adult pedestrian test

Perpendicular adult: Adult walks across road — tests run at 12 and 25 mph

Perpendicular child pedestrian test

Perpendicular child: Child runs into road; parked vehicles obstruct view — tests run at 12 and 25 mph

Parallel adult pedestrian test

Parallel adult: Adult in right lane near edge of road, facing away from traffic — tests run at 25 and 37 mph

As in the vehicle-to-vehicle tests, a GPS system and sensors record data about the vehicle's speed and position, and an on-board camera captures any warnings from the system.

Point system

Vehicles can earn a maximum of 6 points in the vehicle-to-vehicle evaluation. Points are awarded as follows:

  12 mph
test
25 mph
test
Forward
collision
warning
Speed
reduction
(mph)
less than 5 5 to 9 10 or more less than 5 5 to 9 10 to 21 22 or more n/a
Points 0 1 2 0 1 2 3 1

Models with 1 point earn a basic rating. A total of 2 to 4 points qualifies a vehicle for an advanced rating, and 5 to 6 points earn a superior rating.

Some vehicles advertised as having autobrake along with forward collision warning earn only 1 point and a basic rating if the autobrake fails to slow the vehicle enough to earn points in IIHS tests.

The vehicle-to-pedestrian evaluation has its own 6-point scale. For each test scenario, vehicles are awarded points based on the average speed reduction during five test runs.

Speed
reduction
(mph)
5 or less
6 to 11
12 to 17
18 to 23
24 to 29
30 to 36
37
Points 0 0.5 1 1.5 2 2.5 3

An additional point is awarded in the 37 mph parallel scenario to vehicles that warn the driver at least 2.1 seconds before impact.

Total points for the perpendicular scenarios are weighted at 70 percent, while points from the parallel scenario are weighted at 30 percent.

Vehicles with a total score of less than 1 earn no credit for pedestrian crash prevention. Vehicles that earn at least 1 point but less than 3 points earn a basic rating. A score of at least 3 but less than 5 is advanced, while 5 or more is superior.

Headlight evaluation

About half of all fatal crashes in the U.S. occur in the dark, and more than a quarter occur on unlit roads. Headlights have an obvious role to play in preventing nighttime crashes, but not all headlights perform their job equally. Differences in bulb type, headlight technology and even something as basic as how the lights are aimed all affect the amount of useful light supplied.

Headlight technology has been developing rapidly in recent years. LED and high-intensity discharge (HID) lamps have begun to replace the traditional halogen ones. Many automakers offer curve-adaptive headlights, which respond to steering and swivel according to the direction of travel. Many also offer high-beam assist, a feature that can increase the use of high beams by automatically switching between high beams and low beams based on the presence of other vehicles.

At the same time, government regulations about headlights don't guarantee consistency when it comes to the amount of illumination they provide in actual on-road use. This has resulted in large variation in headlight performance. Many vehicles sold today have inadequate headlights, despite the recent strides in lighting technology.

How headlights are tested

IIHS engineers measure the reach of a vehicle's headlights as the vehicle travels straight and on curves. Sensors on the track measure how far from the vehicle the light extends with an intensity of at least 5 lux. A lux is a unit of illuminance, or the amount of light falling on a surface. For comparison, a full moon on a cloudless night illuminates the ground below to about 1 lux.

Both low beams and high beams are measured on five approaches, shown in the graphic below:

  • Straightaway
  • Gradual left curve (800-foot radius)
  • Gradual right curve (800-foot radius)
  • Sharp left curve (500-foot radius)
  • Sharp right curve (500-foot radius)
road curves graphic

On each approach, visibility measurements are taken on the right edge of the roadway. On the curves, measurements also are taken on the left edge of the travel lane. On the straightaway, the second measurement is taken at a point corresponding to the left edge of a two-lane road. This allows the engineers to gauge the illumination on both the right and left side of a straightaway, which are typically quite different. With most headlights, there is a steep drop-off in light on the left side of a straight road in order to prevent glare to oncoming vehicles.

Glare for oncoming vehicles is also measured from low beams in each scenario. Engineers record the percentage by which it exceeds a set threshold.

Headlights are tested as received from the dealer. Although many headlight problems could be resolved by adjusting the aim of the lamps, IIHS doesn't change headlight aim. Few vehicle owners adjust the vertical aim of their headlights, so leaving the aim the way it was set at the factory makes the testing more realistic. Horizontal aim also is important, but in most vehicles it can't be changed after the initial factory setting.

Readings are taken 10 inches from the ground for visibility and 3 feet, 7 inches from the ground for glare.

How ratings are assigned

IIHS engineers compare the results of the testing with a hypothetical ideal headlight system. Using a system of demerits, they apply the visibility and glare measurements to determine the rating.

In this system, the low beams are weighted more heavily than the high beams because they are used more often. The readings on the straightaway are weighted more heavily than those on the curves because crashes are more common on straight sections of road.

A vehicle with no demerits doesn't exceed the glare threshold on any approach and provides illumination to at least 5 lux over the distances shown in the graphic below. Longer visibility distances are required on the straightaway compared with the curves because vehicles tend to travel at higher speeds while going straight. Similarly, greater visibility is required on gradual curves compared with sharp curves.

headlight range graphic

Vehicles equipped with high-beam assist get their low beam demerits reduced. This credit is given only for approaches on which the high beams provide more visibility than the low beams.

LATCH evaluation

Lower Anchors and Tethers for Children (LATCH) is a system of attachment hardware for child restraints. Although child restraints can be installed properly by using vehicle safety belts, LATCH is intended to make correct installation easier. In addition, tether anchors are always needed when installing forward-facing restraints, regardless of whether the lower anchors or the safety belt is used. The top tether keeps the child seat from pitching forward in a crash.

Top tether symbol
Lower anchor symbol

IIHS research has shown that child restraints are more likely to be installed correctly when LATCH is used. However, not all LATCH is easy to use. Parents are more likely to install a restraint correctly when the vehicle hardware meets certain criteria. The Institute established its LATCH rating program in 2015 to encourage manufacturers to design LATCH hardware that meets those ease-of-use criteria.

LATCH hardware
Booster seat installed using LATCH hardware
Lower anchor and tether anchor locations in a typical sedan

Ratings criteria

When vehicles are rated for LATCH ease of use, IIHS employees assign each seating position a rating of good, acceptable, marginal or poor. The following characteristics are noted for the lower anchors:

  • Accessibility of lower anchors. Anchors located at a depth within the seat bight of about ¾ inch or less are considered easy to find. Slightly deeper anchors are OK if they can be easily reached without anything in the way.
  • Force. A special tool representing the lower connector of a child seat is used to measure the attachment force required. The force to attach this tool should be under 40 pounds.
  • Clearance angle. Anything greater than 54 degrees is considered easy to maneuver around.
LATCH testing
LATCH testing
A special tool is used to measure the force required to attach a connector to the lower anchor (left) as well as to measure the anchor's depth within the seat bight. Another tool measures the anchor's clearance angle (right).

When it comes to the tether anchor, evaluators look at the following things:

  • Location. Tether anchors should be on the vehicle's rear deck or on the top 85 percent of the seatback. They shouldn't be at the very bottom of the seatback, under the seat, on the ceiling or on the floor.
  • Confusing hardware. The area where the tether anchor is found doesn't have any other hardware that could be confused for the tether anchor. If other hardware is present, then the tether anchor must have a contrasting label located within 3 inches of it.
Tether anchor
Tether anchor
Tether anchor
Tether anchors can be found in a variety of locations and sometimes are surrounded by other hardware, making installation confusing.

How ratings are assigned

Under existing federal regulations, most vehicles must have at least two rear seating positions with full LATCH hardware and a third with at least a tether anchor. IIHS ratings are based on the best two LATCH positions available in the vehicle's second row.

To earn a good rating, two LATCH positions must meet all five criteria, and a third tether anchor (if required) also must meet both tether criteria. For an acceptable rating, two LATCH positions must each meet at least 2 of the 3 requirements for lower anchors and at least 1 of the 2 tether anchor requirements. If either position fails to meet the tether anchor requirements or meets only one of the lower anchor requirements, then the vehicle is marginal. If even fewer criteria are met, the vehicle is poor.

A rating of good+ is awarded to vehicles that meet the criteria for a good rating and provide additional LATCH-equipped seating positions for maximum flexibility. For a two-row vehicle, that means having a third good or acceptable LATCH seating position. The third position may use either dedicated anchors or borrowed anchors.

For a three-row vehicle to earn good+, it must have one additional good or acceptable LATCH position (without borrowing) and tether anchors in all rear seating positions. The additional tether anchors must meet at least one of the two tether anchor criteria. If the vehicle has a second-row center seating position, it must have good or acceptable LATCH there (borrowing permitted).

What the ratings mean

The LATCH ratings are an indicator of how easy it is to achieve a correct, tight installation of a child restraint in a given vehicle when using the dedicated child restraint attachment hardware. As long as a restraint is properly installed, the LATCH rating doesn't have any bearing on safety. While it's difficult to achieve a good installation in a poor-rated vehicle, it's generally not impossible. In addition, children are just as safe in restraints that have been properly installed with vehicle belts as in restraints that have been properly installed with LATCH.

Verification

Manufacturers are constantly making changes to their vehicles, but time and budget constraints make it impossible for IIHS to test every model every year. To keep our ratings as comprehensive as possible, IIHS engineers work with manufacturers in a process known as verification.

Every year, we determine whether vehicles are changing in any way that could affect their performance in crash tests — for example, if modifications are being made to the structure or to airbags. Engineers gather information about upcoming models from trade journals, auto shows and other sources and confirm their assessment regarding each vehicle with the manufacturer. If there are no changes, the rating from the previous model year gets carried over. If a vehicle has been substantially redesigned, it must be tested again to be included in the ratings.

IIHS doesn't conduct all the tests for every redesigned vehicle itself. If the previous year's model was rated good in the driver-side small overlap front, moderate overlap front or side test, the manufacturer may conduct a verification test in that category. The same is true for vehicles with superior front crash prevention ratings.

In addition, as of 2019, automakers with a proven track record of cooperation with IIHS may conduct verification tests for moderate overlap front and side ratings for any vehicle, even one that isn't a successor model to an earlier good-rated vehicle.

Finally, for any vehicle with a good driver-side small overlap rating, the manufacturer may conduct its own test for the passenger-side rating.

Benefits

Verification allows IIHS to provide consumers with up-to-date information. It ensures that automakers continue to pay attention to frontal and side crash protection as they redesign vehicles. At the same time, it frees Institute engineers to conduct research and develop new tests that in the future will lead to even more improvements in crash protection.

How it works

If a vehicle is eligible for verification, and the manufacturer wants it to receive a rating from IIHS or wants a passenger-side small overlap rating, the company must conduct the test according to Institute parameters.

For crash tests, the company provides video footage, along with measurements of intrusion and injury data from the dummies.

For evaluations of automatic braking for front crash prevention ratings, the manufacturer must provide test footage, as well as information about the vehicle's speed, location and accelerator and brake pedal positions. IIHS engineers analyze this information and assign the rating, just as they would if the test had taken place in the Institute's own Vehicle Research Center.

We only allow verification tests for driver-side small overlap front, passenger-side small overlap front, moderate overlap front, side and front crash prevention ratings. If a vehicle's seats, roof or headlights have changed at all, the Institute conducts a new head restraint, roof strength or headlight evaluation.

To ensure good faith participation, we conduct occasional audit tests in which vehicles are retested in-house to make sure the results don't differ significantly from the manufacturers' tests.

Driver-side small overlap front test — side curtain airbag coverage

IIHS allowed manufacturers to conduct their own tests to show improvement in one specific aspect of the small overlap front rating. After the driver-side small overlap test was introduced, many vehicles received a demerit for insufficient forward coverage from the side curtain airbag. All of these airbags were expected to be modified to improve coverage of the side window area because of a new government safety standard for rollover protection. This new standard, known as FMVSS 226, was phased in by the 2018 model year. IIHS allowed manufacturers to submit small overlap test video for newly compliant vehicles. If IIHS determined that the video demonstrated sufficient airbag coverage, the demerit was removed.