What are the odds of serious injury to a child riding in a rear-mounted bike seat?
Evidence quality 4.25/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 5/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 3/5
- D4 Uncertainty
- 3/5
- D5 Scope
- 5/5
- D6 Prose
- 4/5
- D7 Perception honesty
- 4/5
- D8 Caveat completeness
- 5/5
Lifetime probability · lifetime, activity-specific
1 in 2,500
0.04% lifetime chance
Most people overestimate this.
range 1 in 10,000 to 1 in 500
● your factors — click this risk ▾ to reveal
≈ As likely as
Perceived
A child in a rear-mounted seat sits visibly exposed, head bobbing above the rear wheel, with no seatbelt around the torso and no metal cage between them and the road. Parents imagine the worst-case scenario as a collision with a car, the child flung from a height onto pavement. Some refuse to use mounted seats at all, citing the absence of crash testing standards comparable to car seats. The fear is a composite of motor-vehicle dread, head-injury salience, and the sense that an adult bicycle is an unstable platform for cargo a parent loves.
Rough estimate: ~5-10% chance of serious injury over a typical childhood as passenger
Source: editorial intuition, not polled
Actual
~2,015 ER-treated injuries over 9 years (NEISS-extrapolated, US children riding in bicycle-mounted child seats)
US children <5 transported as passengers on adult-operated bicycles 1990-1998
Show derivation
Powell & Tanz (2000) extrapolated 2,015 mounted-seat injuries (95% CI 988-3,042) over 9 years (1990-1998) from NEISS sentinel-hospital data. The denominator is the dominant uncertainty: no national survey directly counts US children riding as bicycle passengers. Best estimates from cycling participation surveys put the figure at roughly 5-10% of US children under 5, or ~1-2 million per year, giving ~9-18 million cumulative child-passenger-years over the 9-year window. We use a midpoint of 1.5M cumulative passenger-years, yielding ~0.13% per child-passenger-year, or roughly 1 in 750 per year of exposure. Over a typical 3-year passenger window (ages 1-4), the cumulative ER-treated injury probability is approximately 1 in 2,500. The uncertainty band (1 in 500 to 1 in 10,000) reflects the unknown true exposure denominator and the wide CI on the Powell & Tanz numerator.
Caveats: The 1 in 2,500 headline rests on an unobserved denominator. No US survey directl…
The 1 in 2,500 headline rests on an unobserved denominator. No US survey directly counts how many children ride as bicycle passengers, so the per-exposure rate is a midpoint of plausible bounds rather than a measured quantity. The Powell & Tanz data also pre-dates universal helmet messaging, modern integrated spoke guards, and the CPSC small-parts and tip-resistance standards now baked into compliant mounted seats; a current-day NEISS re-analysis would likely show a lower rate. The number does not capture Northern European cargo-bike or Dutch bakfiets configurations, which sit the child lower and forward and have different crash dynamics. Finally, the 9% motor-vehicle share is the figure most parents intuitively fear, but the absolute risk it represents is small: roughly 1 in 28,000 over a 3-year passenger window.
Risks at similar odds
Other risks with roughly the same likelihood — useful for calibration.
Child bike trailer
What are the odds of serious injury to a child riding in a towed bicycle trailer?
Child without restraint
How much more likely is a child to die in a car crash without an appropriate child restraint?
Child window fall
What are the odds of a child being killed or seriously injured by falling from a window or balcony?
Teen road-crash death
How likely is a teenager (15–19) to die in a road-traffic crash during those years?
Recently viewed on this device
Stored locally — clear anytime.
Pick challenger
Powell and Tanz (2000) extrapolated about 2,015 ER-treated injuries over nine years among US children riding in bicycle-mounted child seats — roughly 224 per year nationally. Against a denominator of perhaps a million child passengers per year, that works out to a per-passenger-year rate near 1 in 750, or about 1 in 2,500 over a typical three-year window between when a toddler can sit up unaided and when they outgrow the seat. The single most counterintuitive finding: only 9% of injuries involved a motor vehicle. 72% were falls — the bike tipped in a driveway, a parking lot, or a quiet street.
The perceived/actual gap is shaped by what parents picture. The mental image is a car striking the bike. The reality is a parent stopping at a light, forgetting the rear is now top-heavy, and the whole bike toppling sideways onto a curb. Tip-overs are forgiving to adults, who can step out of a fall. A child strapped above the rear wheel cannot. Head and face injuries accounted for 49% of mounted-seat cases — not because the kinetic energy is high, but because the child’s head is the part that hits first.
Where the number does not apply: the Powell & Tanz cohort pre-dates current CPSC small-parts and spoke-guard standards, so modern compliant seats produce fewer entrapment injuries than the 1990s baseline. The data also captures only ER-treated events, not the larger pool of home-managed bruises and the smaller pool of catastrophic motor-vehicle collisions, which produce few cases but heavy individual severity. Cargo-bike configurations common in the Netherlands and Denmark are absent from this dataset, and the under-12-months use that AAP contraindicates carries materially higher risk than the headline figure.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] Archives of Pediatrics & Adolescent Medicine (Powell & Tanz) — Tykes and bikes: injuries associated with bicycle-towed child trailers and bicycle-mounted child seats
Tykes and bikes: injuries associated with bicycle-towed child trailers and bicycle-mounted child seatsSee all 2 Likelier entries citing this source →
- Statistic
Estimated 2,015 injuries (95% CI 988-3,042) over 9 years to children riding in US bicycle-mounted child seats; 9% involved motor vehicles, 72% were falls; 49% had head/face injuries- Excerpt
“"49 injuries to children during the 9-year study period [were identified]: 6 were associated with bicycle-towed trailers (estimated 322 injuries; 95% CI, 158-486) and 43 were related to bicycle-mounted child seats (estimated 2,015 injuries; 95% CI, 988-3,042). The mean age of injured children was 2.4 years and 51% were male. For trailers, motor vehicle collisions accounted for 33% of injuries and falls 50%; for mounted seats, 9% involved motor vehicles and 72% were falls. Head/face injuries: 83% (trailers) vs 49% (seats)." ”
- Source data from
- 2000-04-01
- Accessed
- 2026-05-31 · archived copy
- Calculation
- Powell & Tanz analyzed NEISS data for 1990-1998 (9 years). The 2,015 figure is the extrapolated US national estimate from observed cases. We use 2,015 / ~1.5M cumulative child-passenger-years (9 years x ~166k US children riding as passengers/year, estimated from cycling participation surveys) for ~0.13% over the 9-year window, or roughly 1 in 750 per child-passenger-year. Over a typical 3-year exposure window, ~1 in 2,500 for ER-treated injury. The denominator is the dominant uncertainty.
-
[2] Pediatric Emergency Care (Tanz & Christoffel) — Tykes on bikes: injuries associated with bicycle-mounted child seats
Tykes on bikes: injuries associated with bicycle-mounted child seats- Statistic
Estimated 4,960 mounted-seat injuries over 11 years (1978-1988); falls accounted for 80%; head 51% and face 21% of injuries; spoke entrapment a notable mechanism- Excerpt
“"We reviewed US Consumer Product Safety Commission (CPSC) data for 1978-1988... There were an estimated 4960 injuries to children during the 11-year period. The peak age of injury was two years. Fifty-five percent of victims were male. Falls accounted for 80% of the estimated injuries. Head (51%) and face (21%) injuries predominated. Twenty-one percent of estimated injuries were mild, 60% were moderate, and 19% were severe." ”
- Source data from
- 1991-10-01
- Accessed
- 2026-05-31 · archived copy
- Calculation
- Tanz & Christoffel 1991 covers 1978-1988 (11-year CPSC dataset). Used for mechanism context: falls dominate (80%), head and face injuries dominate (72% combined), and spoke entrapment was a notable pre-standard mechanism. The era pre-dates modern spoke guards (CFR Title 16 Part 1512 and ASTM F1625), so the injury composition is era-anchored and should not be applied directly to current compliant seats.
-
[3] AAP / HealthyChildren.org — How to Protect Child Passengers on Adult Bikes
How to Protect Child Passengers on Adult BikesSee all 2 Likelier entries citing this source →
- Statistic
AAP guidance: bike-towed trailers preferred over mounted seats; infants <12 months too young to be passengers; helmets required for both child and adult- Excerpt
“"Preferably, children should ride in a bicycle-towed child trailer rather than a bicycle-mounted child seat. Infants younger than 12 months are too young to sit in a rear bike seat or to wear a bicycle helmet." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-05-31 · archived copy
- Calculation
- AAP positions trailers as the safer of the two passenger configurations without publishing a quantitative differential. This is the cleanest authoritative guidance for the comparison-anchor framing against the trailer entry. Also anchors the under-12-months contraindication used in the personal-factor multipliers.