California E-Bike Accident Statistics
Police-reported e-bike collisions in California rose from 2 in 2018 to 1,506 in 2024, totaling 4,035 over that seven-year period, against 58,332 conventional bicycle collisions. Injured e-bike riders skew young: 15.6% were aged 14 or under, against 10.2% of conventional cyclists. But the headline growth figure needs a warning attached. A statewide count of two e-bike collisions in 2018 is not a credible measurement of 2018 reality — e-bikes were already in commercial use, with the three-class framework in law since 2016. That number measures how widely police were using a new data field, not how many crashes happened. Any chart of this series that does not say so is misleading.
California has no reliable e-bike fatality count, no verified helmet-use rate among injured riders, and no year-by-year series for 2019 through 2023. The e-bike field in the state crash database is recent enough that growth partly measures reporting practice. Several widely circulated statistics about California e-bike crashes cannot be traced to a primary source at all. This page states what is verifiable and names what is not, rather than filling the gaps.
The verified figures
Three numbers are solid, traceable to a peer-reviewed analysis of the state crash database, and safe to cite. Everything else on this page is either a documented limitation or an admitted gap.
The source is Austin J, Perkins L, Haines L, Kobayashi L, Berndtson A, Weaver J, Santorelli J, Liepert A, Ignacio R, Burruss S, Maa J, Costantini T, "An exponential increase in e-bike injuries in California," Trauma Surgery & Acute Care Open, published 17 July 2026, DOI 10.1136/tsaco-2025-002099. It is a retrospective analysis of SWITRS — the CHP's Statewide Integrated Traffic Records System — covering police-reported collisions only.
The age pattern is the most robust finding
This is the one comparison in the dataset with clean, like-for-like figures on both sides.
Share of injured riders aged 14 or under, California 2018–2024
Police-reported collisions. Both bars use the same source, the same period and the same definition, which is what makes this the most reliable comparison available.
Injured e-bike riders were younger overall than injured conventional cyclists. Note that this inverts the fatality pattern: conventional-bicycle deaths skew old, with riders aged 55–64 making up 23.4% of bicyclist fatalities in 2023.
View as table
| Rider type | Aged ≤14 | Period | Source |
|---|---|---|---|
| Electric bicycle | 15.6% | 2018–2024 | Austin et al. 2026, SWITRS |
| Conventional bicycle | 10.2% | 2018–2024 | Austin et al. 2026, SWITRS |
Co-author Romeo Ignacio of Rady Children's Health and co-author Jay Doucet of UC San Diego both emphasized the paediatric finding. Doucet: the study found "a dramatic increase in e-bike injuries across California, including a disproportionate number involving children and low rates of helmet use among injured riders."
A separate figure, cited by San Diego County Board of Supervisors Chair Terra Lawson-Remer in September 2026, is that paediatric e-bike trauma cases at Rady Children's Hospital rose from 3 in 2021 to 262 in 2025. The Board voted 4–1 on 16 September 2026 to direct county counsel to pursue litigation against companies marketing illegal electric motorcycles as e-bikes. The same session heard that a state-commissioned study found nearly 88% of electric vehicles at selected middle and high school bike racks were illegal e-motorcycles capable of 40 to 60 mph.
Why the growth figure is not what it appears
The study's headline is that e-bike collisions went from 2 in 2018 to 1,506 in 2024. That is the number every news outlet ran. It is also the number most likely to get this site's credibility destroyed if published without a caveat, so here is the caveat.
A statewide total of two e-bike collisions in 2018 is not a plausible measurement of 2018 reality. E-bikes were already in commercial use in California. The three-class legal framework had been in the Vehicle Code since AB 1096 took effect on 1 January 2016. A count of two measures officer adoption of a new data field, not exposure.
So the curve from 2 to 1,506 is a composite of two things — real growth in e-bike riding, and growth in how consistently police coded e-bikes as e-bikes. Nobody has separated them.
What is missing, and why we are not estimating it. The full text has now been read. The year-by-year counts for 2019 through 2023 exist only inside Figure 1, as a graphic — they are never given as numbers anywhere in the paper. Subtracting the two known endpoints from the seven-year total gives 2,527 for those five years combined, but that is arithmetic, not a data point, and it must not be split across years. This page shows two endpoints and a gap rather than a smooth line through numbers nobody has published.
The date the e-bike code was added to SWITRS and the CHP 555 collision report form could not be established. That single fact would let a reader calibrate the whole series. It is the highest-value outstanding item on this site's verification list. Until it is confirmed, no date is published here.
The two comparisons people get wrong
E-bike versus conventional counts are not a risk ratio
These are raw counts. Dividing one by the other tells you nothing about relative risk, because neither is adjusted for how many miles were ridden on each.
California has no e-bike mileage survey. SWITRS is a crash-count system with no exposure denominator, and no study located estimates California e-bike crashes per mile traveled. This is a genuine data gap rather than an unexamined one: the instrument does not exist.
The direction of the two errors is worth holding onto. Because e-bikes are a small share of the fleet, the raw counts understate per-mile risk. Because they are a fast-growing share, the growth curve overstates the change in per-mile risk. Sources tend to make one error or the other, depending on the argument they are supporting.
Injury severity: a quote, not a statistic
First author John Austin of UC San Diego characterized the mechanism this way: "E-bikes provide valuable transportation and recreational benefits, but their higher speeds can generate crash forces more comparable to motorcycles than traditional bicycles, leading to potentially severe injuries."
That is an attributable expert characterization. It is not a measured rate, and it should be quoted as a quote. The study did find moderate-to-severe injuries more common among e-bike riders than conventional cyclists — but the percentages are not published in any accessible source, and no motorcycle comparison data was obtained. The comparison to motorcycles is an expert's framing of crash physics, not a finding that e-bike injuries match motorcycle injuries.
What does not exist
Naming the gaps is part of the point of this page. Four things people routinely ask for and cite, which cannot be sourced:
A California e-bike fatality count. None. The study reports similar mortality rates between e-bike and conventional cyclists with "a significant increase in e-bike-related deaths in 2024," but no absolute figure for any year appears in any accessible source. If you see a specific California e-bike death count, ask where it came from.
A helmet-use rate among injured riders. The study's authors describe "low rates of helmet use." The number is in their data tables and is not otherwise published. This page does not guess at it.
A head-injury or TBI rate. The paper does not report one. This was previously listed here as a figure that could not be obtained; the full text has since been read, and the rate is absent from the study altogether rather than merely inaccessible. That is a stronger statement and a more useful one — there is no head-injury rate for California e-bike crashes to find.
County or city e-bike breakdowns. This one is worth being precise about, because it is a structural problem rather than a research shortfall. CATSIP and SafeTREC's county datasets do not break out e-bikes — they report bicycle fatalities and serious injuries with no e-bike separation. The Office of Traffic Safety likewise combines e-bikes into a single undifferentiated bicycle category. So there is no official e-bike figure for Los Angeles County, San Diego County, Orange County or anywhere else.
What does exist is local enforcement and crash data that individual cities gathered themselves, usually to justify an ordinance. It is not comparable across cities and not collected to a common standard, but it is the most granular material available:
| Jurisdiction | Figure | Period |
|---|---|---|
| Carlsbad | ~220 bicycle crashes, 63% rider at fault; 12–17 the largest at-fault group; 9 crashes involved riders under 12 | 2022–2025 |
| Oceanside | 188 bicycle collisions, 36% involving e-bikes | Jan 2020 – Jul 2023 |
| Oceanside | 139 citations issued, 11.5% to riders under 12 | Since Sep 2024 |
| Irvine | Riders aged 13–16 accounted for 53% of bicycle collisions | city figure, 2025 |
| Santa Barbara | Nearly 600 citations issued | 2026 |
| Santa Cruz | 16 illegal e-bike impounds, 8 involving riders under 18 | From Aug 2025 |
| Sacramento County | 7 citations, 4 vehicles towed in one targeted American River Parkway operation | Early Jun 2026 |
| La Mesa | 52 bike-related service calls | prior year, cited Apr 2026 |
| Rady Children's Hospital | Paediatric e-bike trauma cases rose from 3 to 262 | 2021 → 2025 |
Two of these figures did real policy work. Oceanside used its citation data — only 11.5% of citations going to under-12 riders — to decline the state's under-12 ban in January 2026. Police Captain Scott Garrett: "The vast majority of our data reflects an older demographic." Carlsbad's crash data pointed the same way, showing 12- to 17-year-olds as the largest at-fault group while its own ordinance restricted under-12s. That mismatch between where the data points and where the legal authority reaches is the central unresolved problem in California e-bike policy.
Statewide bicyclist fatalities, all bicycle types
Useful context, but read the source column carefully — these rows come from two different reporting systems and do not form a comparable series.
| Year | Bicyclist fatalities | Reporting system |
|---|---|---|
| 2017 | 145 | SafeTREC, provisional SWITRS |
| 2018 | 155 | SafeTREC, provisional SWITRS |
| 2020 | 153 | SWITRS via secondary source |
| 2023 | 145 | OTS / SafeTREC, FARS 2023 Annual Report File |
Bicyclist fatalities fell 20.8% from 2022 to 2023. The 2022 absolute figure was not obtained, and back-calculating it from the percentage produces a false precision of roughly 183 — so it is not published here. Figures for 2019, 2021 and 2022 were not obtained.
How to read any California e-bike statistic
Four mechanical reasons two sources citing "SWITRS" will give you different numbers:
Data lag and revision. CHP data entry typically runs 12 to 18 months behind. Provisional data is posted quarterly and counts rise as records are finalised. An early-vintage citation of a given year is systematically lower than a later citation of the same year. This is observable in the sources themselves: one CATSIP dataset is stamped "SWITRS 2020–2024, provisional 2024 update as of 11 December 2025," while an earlier release covered "SWITRS 2016–2020, provisional 2019–2020 as of September 2021." Same database, different vintages, different numbers. Always publish the vintage stamp alongside a SWITRS figure.
Unit mismatch. The source families do not share a unit. Austin et al. count collisions. SafeTREC's facts sheets count fatalities and serious injuries — persons. OTS counts bicyclists "struck and killed." FARS counts fatalities under federal definitions. These are not interchangeable, and SafeTREC's 2025 bicycle fact sheet mixes two systems in one document, taking fatalities from FARS 2023 and serious injuries from provisional SWITRS 2023.
Geocoding attrition. SafeTREC geocodes about 97% of crashes, and only geocoded crashes appear in the TIMS mapping tools — a further 3% loss in any map-derived figure.
Documented underreporting. Caltrans PI-0044 (21 November 2018) establishes that crashes on non-public roadways are "not typically entered in SWITRS," that bicycle collisions are underreported on both public and non-public roadways, and that non-traffic bicyclist hospitalisations substantially exceed traffic-related ones. State Parks' lead law enforcement officer believes park bicycle crashes are "grossly underestimated," since solo crashes are the most common and are rarely reported. PI-0044 gives the direction of the undercount and no magnitude — do not attach a percentage to it.
A note on sources this page does not use
Searches for California e-bike statistics surface a large volume of personal-injury law firm pages carrying crash figures. None were used here, and none are linked.
Part of that is quality control: those pages frequently cite each other, misattribute figures between SWITRS and FARS, and mix units. Part of it is structural — for a reference whose value depends on being citable by AI answer engines, circular citation into competitor content is the fastest way to lose the authority that makes it worth citing. Every figure on this page traces to CHP/SWITRS, Caltrans, OTS, UC Berkeley SafeTREC, FARS, a named city or county record, or the peer-reviewed paper.
Common questions
How many e-bike accidents happen in California each year?
The most recent verified figure is 1,506 police-reported e-bike collisions in 2024, statewide, from a peer-reviewed analysis of SWITRS data published in Trauma Surgery & Acute Care Open in July 2026. The seven-year total for 2018–2024 was 4,035. Both figures count collisions, not injured people, and both count only crashes that generated a police report — which excludes solo crashes, crashes on non-public roadways, and the large share of bicycle crashes that are never reported at all.
How many people die in e-bike accidents in California?
There is no citable figure. The peer-reviewed study found that mortality rates were broadly similar between e-bike and conventional cyclists across 2018–2024, but noted "a significant increase in e-bike-related deaths in 2024." No absolute e-bike death count for any single year appears in any accessible source. Statewide bicyclist fatalities across all bicycle types were 145 in 2023. Anyone quoting a specific California e-bike death count should be asked for the source.
Are e-bikes more dangerous than regular bikes?
The data supports a narrower claim than the question invites. Injured e-bike riders sustained moderate-to-severe injuries more often than conventional cyclists, and skewed younger. But the raw comparison of 4,035 e-bike collisions to 58,332 conventional collisions is not a risk ratio — there is no exposure denominator. California has no e-bike mileage survey, so nobody knows crashes per mile ridden. E-bikes are a small but fast-growing share of the fleet, which means the raw counts understate per-mile risk while the growth curve overstates the trend. Both errors are common.
What percentage of e-bike injuries involve children?
15.6% of injured e-bike riders were aged 14 or under, compared with 10.2% of injured conventional cyclists, over 2018–2024. E-bike riders were younger overall. Note that this inverts the conventional-bicycle pattern: conventional-bicycle fatalities skew old — riders aged 55 to 64 were 23.4% of bicyclist deaths in 2023 — while e-bike injuries skew young. The full age distribution below 14, and the 15–17 band, are not published in any accessible source.
Why do different sources give different California e-bike crash numbers?
Four reasons, all mechanical. Data lag: CHP data entry typically runs 12 to 18 months behind, and provisional counts rise as records are finalised — so an early citation of a given year is systematically lower than a later one. Different units: some sources count collisions, some count injured persons, some count fatalities. Different systems: federal FARS and state SWITRS use different definitions and do not produce comparable series. Geocoding attrition: SafeTREC geocodes about 97% of crashes, and only geocoded crashes appear in the TIMS mapping tools. Always publish the data vintage alongside a SWITRS figure.
Is e-bike crash data in California undercounted?
Yes, and this is documented rather than inferred. Caltrans Preliminary Investigation PI-0044 (2018) found that crashes on non-public roadways are "not typically entered in SWITRS," that "bicycle collisions tend to be underreported on both public and nonpublic roadways," and that non-traffic bicyclist hospitalisations substantially exceed traffic-related ones. The Department of Parks and Recreation's lead law enforcement officer believes park bicycle crashes are "grossly underestimated," noting that solo crashes are the most common and are rarely reported. PI-0044 establishes the direction of the undercount but provides no quantified percentage — do not attach a number to it.
Sources
- Austin J, et al. — An exponential increase in e-bike injuries in California. Trauma Surgery & Acute Care Open 2026;11(3):e002099 (DOI 10.1136/tsaco-2025-002099)
- UC San Diego Today — study release
- TIMS / SafeTREC — SWITRS overview, methodology and data lag
- Caltrans Preliminary Investigation PI-0044 — bicycle crash underreporting
- 2025 SafeTREC Traffic Safety Facts: Bicycle Safety
- CATSIP — pedestrian and bicycle crash data by county
- California Office of Traffic Safety — pedestrian and bicycle safety
- KPBS — San Diego County supervisors approve e-bike lawsuits (16 September 2026)