Child safety

When it comes to crashes, children are much safer than they used to be. The rate of motor vehicle crash deaths per million children younger than 13 is less than a quarter of what it was in 1975. The rate at which children die as passenger vehicle occupants has decreased 65%, while the rates at which they are killed as pedestrians and bicyclists are less than one-tenth of the 1975 rates.

Proper restraint use can reduce crash deaths and injuries even more. Appropriate child safety seats provide significantly more protection in a crash than seat belts alone.

Always seat kids in the rear, and choose the right restraint for your child’s age, height, weight and developmental level.

• All infants and toddlers should ride rear-facing until they reach the height or weight limit of their child restraints, which may not be until age 2 or older.
• Once they outgrow rear-facing restraints, children should ride in a harness-equipped forward-facing child restraint until they reach the height or weight limit of the child restraint. Top tethers should be used whenever a child restraint is installed forward-facing.
• When children outgrow child restraints, they should use belt-positioning booster seats until adult seat belts fit properly.

While the number of child passenger deaths has declined by more than half since 1975, motor vehicle crashes remain a leading cause of death for children. A total of 795 children younger than 13 died in motor vehicle crashes in 2024; almost three-quarters of these deaths were children riding in passenger vehicles.

Proper restraint use can help reduce deaths even more. Although the majority of children ride restrained, 167 children killed in crashes in 2024 were unrestrained, and others were improperly restrained.

Children are safest when they ride in the back seat in the right restraint for their age and size, until they are big enough for adult seat belts to fit properly. That means starting out in a rear-facing restraint before moving up to a forward-facing restraint and then a booster seat.

Restraining children in rear seats instead of front seats reduces fatal injury risk by about three-quarters for children up to age 3, and almost half for children ages 4 to 8 (Durbin et al., 2015). In the front seat, children, particularly infants in rear-facing child restraints, may be at risk of injury or death from an inflating front airbag.

Any restraint is better than none at all, but an appropriate child safety seat provides the best protection in a crash until children are large enough for adult seat belts to fit properly, usually when a child is about 4 feet 9 inches tall.

Appropriate child safety seats provide significantly more protection in a crash than seat belts alone. Harness-based child restraints reduce fatal injuries by 58%-71% for infants (younger than 1) and by 54%-59% for 1-4-year-olds compared with no restraint (NHTSA, 2009). In comparison, adult seat belts reduce the risk of death in a crash among 1-4-year-olds by 36%.

Children 2-6 years old in child safety seats (including child restraints and belt-positioning booster seats) are about 28% less likely to be fatally injured than those using seat belts alone (Elliot et al., 2006). Children ages 4 to 8 using belt-positioning boosters are 45% less likely to be injured than children using belts alone (Arbogast et al., 2009).

The National Highway Traffic Safety Administration (NHTSA) estimates that child safety seats saved 408 lives in 2019 and an estimated 13,217 lives from 1968 to 2019 (Kahane & Simons, 2024).

The best seat for a child depends on the child’s height, weight, age and developmental level. The best seat is also one that the care provider can use correctly on every trip.

Rear-facing child restraint: A rear-facing child restraint in the back seat is best for babies and young children. Twenty-five states and the District of Columbia require children to be rear-facing until at least age 2. Never put a rear-facing restraint in the front seat of a vehicle with an active frontal airbag. For the best protection, young children should sit rear-facing as long as possible, until they grow too big for the restraint’s height or weight limit.

After outgrowing infant-only child restraints like the one shown above — around age 1 — children should be placed in a rear-facing convertible child restraint. These seats can be used rear-facing until the child reaches the maximum height and weight limit for rear-facing use. The limit is typically around 40 pounds and 40 inches, although many seats available today offer higher rear-facing height and weight limits.

Forward-facing child restraint: Forward-facing harness-equipped restraints provide excellent protection when used properly. Children can ride in these seats from about age 2 to 7 years. Children should ride in forward-facing restraints for as long as possible, up to the height or weight limit of the child restraint. Some forward-facing seats have weight limits as high as 90 pounds.

Booster seat: Kids too big for child restraints should use belt-positioning booster seats until adult belts fit right. For some kids, that's not until age 12.

With or without a booster, the shoulder belt should fit snug across the center of the shoulder and not across the neck or face or slipping off the shoulder. The lap belt should lie flat across the upper thighs and not on the tummy.

A national observation survey found that 88% of child passengers under age 13 were restrained in 2023 (Werth, 2025). Ninety-six percent of children from birth to 12 months were restrained, compared with 94% of children ages 1 to 3, 87% of children ages 4 to 7 and 84% of children ages 8 to 12. 

While most child passengers are restrained, they aren’t always in the proper restraint for their height and weight. In 2023, 93% of children under 20 pounds were riding in rear-facing restraints, while 4% were prematurely placed in forward-facing child restraints and 3% rode unrestrained.

Among children from 20 to 40 pounds, 18% were still in rear-facing seats, while 60% were in forward-facing child restraints. The others were in booster seats (11%), were using seat belts only (5%) or were unrestrained (7%). Premature graduation to seat belts was common among older children. Among children 37 to 53 inches tall, 23% were using belts only, rather than booster seats. An additional 12% were unrestrained. Among children 54 to 56 inches tall, marginally below NHTSA’s 57-inch benchmark, 69% were using belts only and an additional 15% were unrestrained.

All 50 states and the District of Columbia have child safety seat laws requiring children under a certain age to travel in federally approved child restraint devices. The age at which belts can be used instead of child restraints differs by state.

Young children usually are covered by child safety seat laws, while seat belt laws cover older children and adults. Ideally, all infants and children in all motor vehicles should be covered by seat belt laws, child safety seat laws or both. But differences in the wording of state laws mean some children aren't being covered by either law.

In recent years, lawmakers have eliminated these gaps in most states. All children younger than 16 in 47 states and the District of Columbia are covered by one or both laws.

Child safety seats reduce child deaths and injuries, but they can be difficult to install and are commonly misused.

In 2011, NHTSA observed children riding in more than 4,000 vehicles and found more than 46% of the children were riding in child safety seats that were being misused in a manner likely to reduce protection in a crash. Misuse rates were highest (61%) for forward-facing child restraints. The most common types of child restraint misuse were loose attachment to the vehicle seat and loose harness straps (NHTSA, 2015).

During observations at a hospital in 2013-14, 91% of child restraint installations for newborns ready for discharge had one or more serious misuses (Hoffman et al., 2016). The most common were loose harness straps, unlocked seatbelt retractors, loose child restraint attachment to the vehicle, and incorrect recline angle.

Top tethers are recommended for use with all forward-facing child restraints but many parents do not use them. An IIHS observational survey conducted in 2012, top tethers were used in only 56% of forward-facing restraints (Eichelberger et al., 2014). Parents who were not using the top tether most often did not know the tether was on the child restraint (22%) or did not know how to use it (15%). Ten percent were unsure of where to attach it in the vehicle. In an IIHS observational study conducted in 2010, top tether use was particularly low for older vehicles and pickups (Jermakian et al., 2011).

Other IIHS research found that parents are more likely to use the top tether if the tether anchor in the vehicle is easy to find, such as on the rear shelf of passenger cars or the middle of the seatback in SUVs and minivans, and if there is no other hardware that could be confused with a tether anchor (Cicchino & Jermakian, 2015; Jermakian et al., 2014).

Misuse of belt-positioning boosters also is common. In the 2011 NHTSA observational survey, the misuse rates were 24% for backless boosters and 16% for highback boosters. The most common errors involved a lap belt positioned too high so that it crossed the abdomen or rib cage instead of the upper thighs and a shoulder belt positioned behind the back or under the child's arm (NHTSA, 2015).

A backover crash occurs when a vehicle backs into a person such as a pedestrian or bicyclist, often when exiting a driveway or parking spot. Young children and older people are most likely to be killed in a backover.

The extent of the backover problem is difficult to determine because government databases generally record only crashes on public roads, but most backover crashes occur in driveways and parking lots. However, NHTSA's Non-Traffic Surveillance (NTS) crash database of nontraffic events resulting in injuries and deaths can be used to calculate a national annual estimate. Based on 2017-21 NTS data and data about on-road crashes, NHTSA estimates that 263 deaths and about 10,000 injuries occur annually in backover crashes (NHTSA, 2024).

NHTSA estimates that 39% of backover fatalities occur in residential spaces such as driveways and the parking lots of apartment and townhouse complexes (NHTSA, 2008). Nonresidential parking lots account for only 17% of backover fatalities, but 52% of backover injuries. 

An analysis of driveway backovers involving children in Utah in 1998-2003 found that children were more likely to be injured by a pickup truck, minivan or SUV than a car, relative to the number of registered vehicles of each type, although the difference between SUVs and cars was not significant (Pinkney et al., 2006). Larger vehicles like SUVs and pickup trucks typically have bigger blind zones than cars (Mazzae & Garrott, 2008; Mazzae & Barickman, 2009), in large part because they sit higher off the ground, making it more difficult for drivers to see children and smaller objects near the rear of the vehicle. 

Consumer Reports has measured distances behind the rear of a vehicle that a driver cannot see and has found that a 5-foot-8-inch-tall driver in an average midsize SUV cannot see an area extending 18 feet behind the vehicle, compared with 13 feet for an average midsize sedan (Consumer Reports, 2014). NHTSA measurements of rear visibility also have found that blind zones for shorter drivers are typically much bigger.

Rearview cameras greatly improve visibility behind a vehicle. An IIHS study found they reduce the zone in which a toddler can’t be spotted by an average of about 90% (Kidd & Brethwaite, 2014).

Rearview cameras have been essentially required on all new vehicles under 10,000 pounds since May 2018 (Office of the Federal Register, 2014).

Frontover crashes, or low-speed crashes in which a driver hits someone in their front blind zone, are a growing concern for child safety. The forward blind zones of several popular SUVs and a popular pickup grew substantially between 1997 and 2023 (Epstein et al., 2025). These large blind zones make it difficult to see small children in front of the vehicle. Information on the extent of the problem is even more limited for frontover crashes than for backovers.