ROR Labs cover: 0 seen unrestrained. A national survey watching real cars found no infants under one riding without a seat. A quarter of the babies who died in crashes were.
|

Nobody Skips the Car Seat. The Deaths Are in the Exceptions.

Disclosure: this article contains affiliate links, marked (paid link). If you buy through one we may earn a commission, at no extra cost to you. As an Amazon Associate I earn from qualifying purchases. It costs you nothing and it does not change what we recommend.

Key takeaways · 8 min read

  • Child restraints cut fatal injury risk by 71% for infants and 54% for toddlers in passenger cars.
  • In 2024, 550 children under 13 died as vehicle occupants; 167 were unrestrained.
  • Observation surveys find near-universal use, and the deaths cluster in the exceptions and the misuse.
  • Rear-facing as long as the seat allows, harness to at least 4, booster until the belt fits at about 4 ft 9 in.

One gap in the child car seat data says almost everything about this subject. In a national observation survey, researchers watching real cars found no infants under one year riding unrestrained. Yet in the fatality records for the following year, a quarter of the babies who died in crashes were unrestrained.

Almost everyone uses a car seat almost all of the time, and the deaths are concentrated in the exceptions: the quick trip, the other car, the borrowed lift, the day the seat was in the other vehicle. A second, quieter problem sits beneath that: seats that are used but used wrong.

A car on a driveway in morning light with the rear door open and a child seat visible inside.
The trip that goes wrong is the one nobody planned.

How much a car seat is actually worth

Reduction in risk of fatal injury from a child restraint

Versus no restraint. NHTSA / National Safety Council figures, 2024 data.

Passenger carsLight trucks
71%
58%
54%
59%
infants under 1toddlers 1–4
A 71% reduction for infants in passenger cars. There is very little else a parent can do that moves a number this far.

Source: National Safety Council, Injury Facts — Child Restraint, drawing on NHTSA data (2024).

Child occupant deaths, and how many were unrestrained

Children under 13, United States, 2024.

550child occupants under 13 died in traffic crashes
167of them were unrestrained — about 30%, with many others inadequately restrained
“Inadequately restrained” is the category doing the quiet work: the right seat used incorrectly, or the wrong stage for the child’s size.

Source: National Safety Council, Injury Facts — Child Restraint (2024 data).

The stages, and the pressure to skip them

Every transition in this sequence trades some protection for convenience or for a child’s sense of being grown up. Each is worth delaying as long as the seat allows.

The four stages, per the American Academy of Pediatrics

The rule throughout is the same: stay in each stage until the child outgrows the manufacturer’s height or weight limit, not until a birthday.

StageHow longWhy it matters
Rear-facingAs long as possible — to the seat’s top height or weight limit. Most convertible seats allow past age 2.Spreads crash force across the whole back and supports the head and neck. A young child’s head is disproportionately heavy and the neck cannot hold it.
Forward-facing harnessAfter outgrowing rear-facing. Keep the harness at least to age 4, and longer if the seat allows.A five-point harness spreads load across the strongest parts of the body and keeps the child in position.
Booster seatUntil the adult belt fits properly — typically about 4 ft 9 in, usually between 8 and 12 years old.A booster does one job: it raises the child so the adult belt crosses the shoulder and the hip bones, not the neck and the abdomen.
Seat belt aloneWhen the belt genuinely fits. Most children are 10 to 12 before that is true.A lap belt riding on a soft abdomen instead of the pelvis is a recognised injury pattern in crashes.
And through all four stages: every child under 13 belongs in the back seat.

Source: American Academy of Pediatrics, car safety seat guidance (HealthyChildren.org).

Most of the argument happens at the rear-facing stage, usually because a toddler’s legs look cramped. Bent legs against the seat back are no safety problem, and children sit cross-legged comfortably in ways adults find alarming. Leg fractures in rear-facing children are rare. Head and neck injuries in forward-facing toddlers are not.

The five mistakes that matter

A child car seat seen from the front with its harness straps hanging loose.
Loose is invisible until the moment it is not.

Most fire stations and many hospitals will check a seat for free, because installation errors are common. These are the ones that recur.

What to check tonight, in two minutes

Each of these is a documented common error.

1. HARNESS TOO LOOSEThe pinch test: at the collarbone, you should not be able to pinch a horizontal fold of webbing between finger and thumb. If you can, tighten it.
2. CHEST CLIP TOO LOWIt belongs at armpit level. Down on the belly it lets the shoulder straps slip off in a crash, which is the whole failure mode it exists to prevent.
3. BULKY COATSA padded coat compresses instantly in a crash, leaving several centimetres of slack. Harness the child first, then put the coat on backwards over the top, or use a blanket.
4. SEAT MOVES TOO MUCHGrip at the belt path and pull side to side and front to back. Under an inch of movement. Not at the top of the seat — the top is expected to move.
5. BELT AND LATCH TOGETHERUse one or the other, not both, unless the manufacturer explicitly permits it. And LATCH anchors have a combined weight limit — past it, you switch to the seat belt.

Sources: NHTSA child passenger safety guidance; American Academy of Pediatrics installation guidance.

Harness and clip position

The two adjustments most often wrong.

CORRECT clip at armpit level straps stay on the shoulders no slack at the collarbone WRONG clip on the belly straps spread apart — the child can come out
The chest clip is a positioning device, not a restraint. Its only job is holding the two shoulder straps over the shoulders, and it can only do that from armpit height.

Source: NHTSA and AAP child passenger safety guidance.

Buying a seat

The expensive seat and the cheap seat both pass the same crash standard. Fit and correct use are what vary.

CONVERTIBLE CAR SEATRear-facing then forward-facing in one seat. Look for a high rear-facing height and weight limit — that number is what lets you stay in the safest stage longest.
Browse on Amazon →
HIGH-BACK BOOSTERFor the booster years. The high back positions the shoulder belt correctly and gives head support in cars without high seat backs.
Browse on Amazon →
MIRROR FOR REAR-FACINGLets you see a rear-facing baby without turning around, which is the actual safety benefit — a driver looking over their shoulder is not looking at the road.
Browse on Amazon →

Two categories we will not link. Second-hand car seats — a seat involved in a crash can be structurally compromised invisibly, plastic degrades, and you cannot verify a stranger’s history or recall status. And aftermarket strap covers, head supports, seat protectors and inserts that did not come with the seat: they were not crash-tested with it, they can introduce slack or change how the harness sits, and many manufacturers state that using them voids the seat’s certification. The seat was tested as it was sold.

A child car seat standing on a hallway floor, taken out of one car and not yet fitted in another.
It only protects the car it is fitted in.

The four stages and the checks NHTSA prints

This section reports what NHTSA, the CDC and one paediatricians’ group say about child passenger seats. It is general information; your seat’s manual and vehicle manual decide the details.

Stage and what the pages say
StageWhat the pages say
Rear-facingNHTSA prints birth to 12 months and 1 to 3 years, and says to keep a child rear-facing as long as possible, up to the seat’s top height or weight limit; children under 1 should always ride rear-facing. The CDC says rear-facing from birth until about age 2 to 4. The AAP (a nonprofit professional body) says most convertible seats allow about 2 years or more.
Forward-facing with harnessNHTSA: 1 to 3 years and 4 to 7 years, using the harness and tether. The CDC: until the seat’s limit and at least age 5.
BoosterNHTSA: 4 to 7 years and 8 to 12 years, until the adult belt fits. The CDC: usually until age 9 to 12. The AAP says about 4 feet 9 inches and 8 to 12 years.
Adult seat beltNHTSA: 8 to 12 years, in the back seat. The belt fits when the child sits without slouching with the back against the seat, knees bend over the seat edge and feet are flat; the lap belt sits low on the upper thighs and the shoulder belt on the shoulder and chest, not the neck or face. The CDC says back seat until age 13.
After a crashNHTSA: a crash is minor only if the car could be driven away, the door nearest the seat is undamaged, no one was injured, airbags did not deploy and the seat shows no visible damage. If any of these fails, replace the seat. Never use a seat from a moderate or severe crash.
Lower anchors (LATCH)Limits are set by the car and seat makers, so read the label. If there is none, NHTSA gives: lower anchor weight limit = 65 lbs minus the weight of the car seat. Past that, install with the seat belt.
Used or old seatsNHTSA’s checklist says to use the manufacture date and model number to check recalls and whether the seat is too old. It gives no number of years.
Getting it checkedFree checks by certified child passenger safety technicians, some virtual; NHTSA points to its car seat page and to Safe Kids Worldwide.

On the numbers, NHTSA says 677 children aged 14 and under died in passenger vehicles in 2024 and that car seats reduce the risk of fatal injury to infants by 71 percent. The CDC says car seats reduce crash injury risk by 71 to 82 percent compared with a seat belt alone, and booster seats cut serious injury risk by about 45 percent for ages 4 to 8.

What we could not confirm: a numbered five-step belt test on any NHTSA or CDC page (they give the criteria above), a fixed expiry period in years, a phone number for finding an inspection station, and dates for several NHTSA pages, which show none. NHTSA’s age bands overlap, so treat ages as rough and the seat’s own limits as binding.

What to do

Steps sorted by when they matter. Each one says how strong the evidence is: Measured Official advice Untested

Before

  • Do the pinch test: at the collarbone you should not be able to pinch a horizontal fold of harness webbing. Official advice see Sources
  • Put the chest clip at armpit level, not on the belly. Official advice see Sources
  • Grip the seat at the belt path and pull: under an inch of movement side to side and front to back. Official advice see Sources
  • Use the seat belt or LATCH, not both, unless the manufacturer explicitly permits it. LATCH anchors have a weight limit. Official advice see Sources
  • Many fire stations and hospitals will check a seat for free. Official advice see Sources

If it happens

  • No bulky coat under the harness. Harness the child first, then put the coat on backwards over the top, or use a blanket. Official advice see Sources

What not to do

  • Do not use a second-hand seat or add-ons that did not come with the seat. Crash damage can be invisible, and inserts were not crash-tested with it. Untested see Sources

Questions people ask

Do car seats really expire?

Manufacturers print an expiry date, typically six to ten years from manufacture, stamped on the shell rather than the box. The reasons are plastic degradation from temperature cycling in a car, standards moving on, and labels and instructions becoming illegible. The date is also convenient for manufacturers, which is a fair objection, but the temperature cycling is real, and a seat that has spent eight summers in a hot car is not the seat that was tested.

Is the middle of the back seat best?

If you can get a solid installation there, yes, since it is furthest from any impact point. But many cars have a raised or narrow middle seat, or no lower anchors for it. A tight install in an outboard seat beats a loose one in the middle. Position is a smaller factor than installation quality.

Does an expensive seat protect better?

Every seat sold legally meets the same federal crash standard. A higher price buys easier installation, better fabrics, longer height and weight limits, and side-impact features beyond the minimum. Easier installation is a real benefit given how common misuse is, yet a cheap seat installed correctly outperforms an expensive one installed badly.

My child insists they are too big for a booster.

Use the test rather than the argument. Sit them on the vehicle seat and check four things: do their knees bend at the edge, does the lap belt sit low across the hip bones, does the shoulder belt cross mid-shoulder, and can they stay that way for the whole trip? If any answer is no, the belt does not fit yet. Showing a child the actual test tends to work better than an instruction, and backless boosters are far less conspicuous to a self-conscious eight-year-old.

The short version

  • Child restraints cut fatal injury risk by 71% for infants and 54% for toddlers in passenger cars.
  • In 2024, 550 children under 13 died as vehicle occupants; 167 were unrestrained.
  • Observation surveys find near-universal use, and the deaths cluster in the exceptions and the misuse.
  • Rear-facing as long as the seat allows, harness to at least 4, booster until the belt fits at about 4 ft 9 in.
  • Every child under 13 in the back seat.
  • Two-minute check: harness passes the pinch test, chest clip at armpit level, no bulky coats, under an inch of movement at the belt path.
  • Free car seat checks are available at many fire stations and hospitals.
  • Avoid second-hand seats and aftermarket inserts that did not come with the seat.

On the links above: some are affiliate links, marked (paid link). If you buy through one we may earn a commission at no additional cost to you. As an Amazon Associate I earn from qualifying purchases. We link to product searches rather than specific items so recommendations do not break as models change, and we say plainly when we are choosing not to link something. Full policy: Affiliate Disclosure.

Sources

  • National Safety Council. Injury Facts — Child Restraint (2024 data, drawing on NHTSA fatality and observation surveys).
  • American Academy of Pediatrics. Car safety seat guidance for families, HealthyChildren.org.
  • National Highway Traffic Safety Administration. Car seats and booster seats guidance; Evaluation of Child Restraint System Effectiveness.
  • NHTSA. Car Seats and Booster Seats. nhtsa.gov; Car seat recommendations for children by age and size (PDF), nhtsa.gov; Car seat use after a crash, nhtsa.gov; Used car seat safety checklist, nhtsa.gov; Seat belts, nhtsa.gov; Right seat, nhtsa.gov.
  • CDC. Child passenger safety: prevention, updated Apr. 23, 2026. cdc.gov.
  • American Academy of Pediatrics (HealthyChildren.org; nonprofit). Rear-facing car seats (Feb. 28, 2021) healthychildren.org; Booster seats healthychildren.org.

Similar Posts