Garage door injuries among children: what the federal data shows

What did we find?

We could not query NEISS, and we will not borrow its figures secondhand. NEISS is the federal injury database run by the Consumer Product Safety Commission (CPSC). The cpsc.gov site returned HTTP 403 (access blocked) to every request we made. So this page has no HyreGarage-computed national rate of child injuries from garage doors.

A peer-reviewed study did the work directly. Kriel and colleagues, in Pediatrics in October 1996, found that "at least 85 children have had permanent brain injury or have died since 1974" from automatic garage door opener accidents in the United States.

The same study found the gap the federal rule exists to close. The authors tested 50 openers themselves. 88 percent reversed when they met a block of wood. But 40 percent failed to reverse when coming down on a child-sized CPR mannequin.

This finding came after federal rulemaking. CPSC’s own 1991 rulemaking record counted 48 child entrapment deaths since 1982, before the entrapment-protection standard took effect. Five years later, the Kriel study found openers already in American homes still failing a child-size test. Our garage door injury study for all ages covers that record in full.

How does this connect to the federal safety standard?

The study shows why the standard was needed and why it took time to work. Our garage door injuries study found, in CPSC’s own 1991 rulemaking record, 48 child entrapment deaths since 1982. Those deaths came before the entrapment-protection standard took effect.

That standard requires an automatic opener to reverse when it meets an obstruction. The 48 deaths were the record’s own reason for the rule.

The Kriel study came out in 1996, after that rulemaking. Its authors tested openers already in American homes. 40 percent of the 50 they tested failed to reverse on a body-like obstruction. Yet 88 percent passed a cruder test with a block of wood.

This does not tell us how many of today’s standard-compliant openers would fail. The standard itself requires tougher reversal testing than the informal wood-block or mannequin methods the study used. It does show, with a tested sample, that "reverses against something" and "reverses against a child" are different claims.

HyreGarage analysis: read together, the two figures describe a before and after that one rule did not fully settle. The 48 deaths justified the requirement. The 85-child figure and the 40 percent failure rate came five years after the rule began phasing in.

That suggests the problem lasted in older installed openers for some time. A rule existing is one thing. Every affected device complying with it is another. See our study of openers built before the federal standard for that history.

What would NEISS tell us, and why couldn’t we ask it?

NEISS would give a weighted national estimate, but it was blocked. The National Electronic Injury Surveillance System is the federal dataset built for this question. It estimates emergency-room injuries nationwide. CPSC’s public tools let you search it by product code and by patient age band.

Our garage-door-injuries page already found that cpsc.gov blocked us with HTTP 403. That covered the NEISS query tool, its data archive and the site root, on the retrieval date. The same block applies to this narrower question about children.

We found no workaround. We are not presenting a number from elsewhere as if it were our own NEISS calculation.

The one NEISS-based figure we have comes from a peer-reviewed study, not our own query. Shapovalov and colleagues (American Journal of Emergency Medicine, January 2026) studied NEISS records for automatic doors in general over 2015-2024.

They found children made up 11 percent of an estimated 38,720 weighted emergency department visits across the decade. But "automatic doors" includes supermarket entrances and powered sliding doors as well as garage doors.

The abstract also does not say which NEISS product codes were used. So no garage-door-only share for children can be pulled from it.

What could we not find?

A current national estimate of garage door injuries to children. This is the number a reader would most want. NEISS, the dataset built to produce it, could not be queried from our environment. We report the published research instead of guessing a figure we could not check.

Breakdowns by age band, cause and body part. Parents often want these, split by child age band from NEISS. Without NEISS access we could not produce them. The Kriel abstract does not report them at that level of detail either.

Whether the 40 percent failure rate applies to today’s openers. The Kriel study tested openers as of 1996, some built before the federal standard took effect. We make no claim about how often current compliant openers fail.

Those openers must pass a tougher test than either the wood-block or mannequin method the study used informally.

A death or injury count since 1996. We found no newer peer-reviewed study updating the 85-child figure for garage door openers specifically, as opposed to automatic doors in general.

How the 85 cases split by outcome or cause. The abstract gives the combined figure (death or permanent brain injury) and the general entrapment cause. It does not tally each case. We did not retrieve the full text beyond the abstract.