ATV Accident Statistics: What the Federal Data Shows and What It Means for Injured Riders

Purpose

This resource shows, in one place and in the report's own words, what the federal government records about deaths and injuries involving off-highway vehicles. For a family trying to make sense of a crash, that matters because the report's categories and wording shape what the numbers can and cannot say about any one event.

An off-highway vehicle (OHV), as the report uses the term, is a vehicle with more than two wheels built for off-road use. The report sorts OHVs into three categories:

• An all-terrain vehicle (ATV) is an off-road motorized vehicle with three or more low-pressure tires, a straddle seat for the operator and handlebars for steering control.
• A recreational off-highway vehicle (ROV) has four or more tires, side-by-side seating and automotive-type controls (a steering wheel and pedals).
• A utility terrain vehicle (UTV) shares those features with the ROV. The report counts the two separately.

Every figure on this page comes from one source: the 2025 Report of Deaths and Injuries Involving Off-Highway Vehicles with More than Two Wheels, which presents information collected by CPSC staff. Its death figures cover 2020 through 2022, which staff considers the most recent complete, or nearly complete, three-year period. Its emergency department injury estimates cover 2020 through 2024. The figures on this page are for all ages combined, so families looking into a child's injury will see the national picture rather than a separate count for young riders.

The wording matters. Death figures are deaths "CPSC staff is aware of," not a complete census. Injury figures are "estimated." Each figure is "associated with" OHVs, a phrase that links an incident to a vehicle type without saying what caused anything. A useful first question for anyone reading these numbers is which vehicle category their own crash falls into, since the report counts each one separately.

What the Federal OHV Data Covers — and What It Does Not

A national statistic describes a population of incidents. It does not establish what caused any individual crash, and it is not a finding about any company, product, park or operator.

Families often already hold records that speak to their own event: an incident or police report, emergency room and hospital records, photos and video of the scene and the vehicle, the vehicle itself, purchase or rental paperwork, and the names of witnesses. The federal data cannot replace any of these. It gives context: the kind of incident the government tracks, the vehicle categories it separates, and the mechanisms it sees most often.

What the data leaves unresolved is what happened in one crash, who controlled what, and which state's law applies. Those questions turn on the facts of a single event, which is why the records a family keeps tend to carry more weight than any national figure.

Methodology

How the report gathers its numbers affects how much a reader can take from them. A count of reported deaths and an estimate drawn from a hospital sample answer different questions, and someone comparing their own crash to the data may find it helpful to know which kind of figure they are looking at.

How This Tool Works and What Its Results Mean

The statistics tool on this page draws only from the CPSC's 2025 OHV Annual Report, dated May 2026. It does not blend in other data sets, news reports or estimates of its own. Each figure is displayed verbatim, with its reporting period, the population it describes and the report's own qualifier.

If a death is part of your family's situation, it helps to know where the death count comes from. Those figures come from the CPSC's Consumer Product Safety Risk Management System, which includes death certificates from the 50 states and the territories, among other report types. The report describes these data as anecdotal, gives the counts as reported to CPSC as of September 2025, and states that additional incidents may be recognized in the future. A recent death may therefore not yet appear in the totals.

If someone was treated in an emergency room, the injury figures are the relevant ones. They are national estimates drawn from NEISS, the National Electronic Injury Surveillance System, a sample of hospital emergency departments. The estimates come from NEISS cases coded under any of five product codes used for ATVs, ROVs and UTVs, and possibly other unspecified off-highway vehicles. The report states that in 2024 the CPSC revised the NEISS sample so it remains nationally representative. Adjusted weights were applied to earlier years, so estimates for 2023 and earlier differ from those in prior reports.

People treated for OHV injuries may also have been asked about their crash. Starting in January 2022, CPSC staff invited people treated for OHV injuries to take part in a follow-up interview or online survey, mainly to better understand how injuries are actually split among ATVs, ROVs and UTVs. All of its questions were optional.

A result from the tool is informational. It does not determine fault, liability, a deadline or what a case is worth. Those depend on the facts of the individual incident and the law of the state involved. What a reader may take from a result is context for questions to raise with an attorney, not an answer to them.

The tool also reflects one legal distinction. The federal ATV standard applies to ATVs as its own text defines them. Under 15 U.S.C. § 2089, an ATV is a motorized off-highway vehicle designed to travel on three or four wheels, with a seat straddled by the operator and handlebars for steering control. The definition excludes prototypes intended only for research and development unless they are offered for sale. The implementing regulations at 16 C.F.R. § 1420.1 and 16 C.F.R. § 1420.3 set requirements for ATVs. Section 1420.3 requires new four-wheel ATVs manufactured on or after January 1, 2025, to meet the 2023 ANSI/SVIA standard, except for its owner's manual section. Earlier vehicles follow the 2017 version. That standard is written for ATVs as defined. It does not by its own terms reach ROVs or UTVs. For someone hurt in a side-by-side, that means the vehicle type and its manufacture date are among the details that shape which standards may be relevant.

How to Read the Federal OHV Data

The report's figures read most accurately when each is checked against the same points:

1. The reporting period. Deaths cover 2020 through 2022; injuries cover 2020 through 2024.
2. The kind of figure. Deaths are reported counts staff is aware of; injuries are estimates from a sample.
3. The vehicle category. ATV, ROV and UTV figures are separate, and labels can be unreliable.
4. The qualifier. "Associated with" links an incident to a vehicle type, not to a cause.
5. The population. Injury figures count only emergency department-treated injuries; the rollover and ejection percentages come from a study of non-fatal injuries.
6. The gap to one crash. No figure decides fault, defect or legal responsibility in an individual incident.

Interpretation

The figures below are the ones the tool displays, each with its period and qualifier. For a reader, each one answers a narrow question about a group of incidents. The questions that matter to a claim, such as who was responsible and what happened in one crash, sit outside them, and the notes under each figure point to what a person may weigh next.

Deaths Associated with OHVs: The Three-Year Figures

The report states: "CPSC staff is aware of 2,685 deaths associated with OHVs that resulted from 2,606 incidents between 2020 through 2022." Because there were fewer incidents than deaths, some incidents involved more than one death. In rare cases a person died in a later year than the year of the incident.

By vehicle type, the report states that "CPSC staff associates 1,759 with all-terrain vehicles (ATVs), 707 with recreational off-highway vehicles (ROVs), and 53 with utility terrain vehicles (UTVs)." For the remaining 166 deaths, staff concluded the vehicle was either an ROV or a UTV but did not know which. ATVs account for the largest share of the three-year total.

For a grieving family, those category lines can matter more than the totals, because the legal theories that may apply, and the standards a vehicle was built to, can differ by vehicle type. Confirming the exact vehicle involved, through purchase or rental paperwork or the vehicle itself, is one of the questions a family may weigh. The report gives a three-year total; this page does not divide it into an annual figure. None of these deaths is attributed here to any company, product or operator.

Emergency-Room Injuries Over Five Years

The report states: "Over the 5-year period from 2020 through 2024, there were an estimated 578,700 emergency department-treated injuries in the United States that were associated with OHVs." It continues: "This corresponds to an annual average of around 115,700 emergency department-treated injuries over the period." The report also estimates the yearly rate ranged from around 32 to 38 per 100,000 people from 2020 to 2024.

If you were taken to an emergency room after a crash, your visit is the kind of event NEISS is built to sample. "Estimated" reflects that NEISS is a probability sample, not a count of every case. The report gives coefficients of variation, a measure of statistical uncertainty, of around 19 to 31 percent for its injury estimates broken out by vehicle type.

The figure counts only emergency department-treated injuries. Injuries treated elsewhere, such as in urgent care or a doctor's office, or not treated at all, are outside it, so the figure is better read as a floor than a ceiling. What matters for an individual is the record of their own treatment: the emergency room and hospital records, and any later care, are what describe a specific injury. These estimates are not findings about any product, company or operator.

The Two Hazards Behind Most Fatalities — and Why They Matter If You Were Hurt

The report states: "The most common fatality hazards associated with OHV-related deaths were overturns and collisions with another vehicle or a stationary object (like a tree)." An overturn, or rollover, is when the vehicle tips onto its side or roof. An ejection is when a rider or passenger is thrown from the vehicle.

Staff's review of 2020–2022 deaths found overturning was a primary hazard in about 42 percent of fatalities and collisions in around 27 percent. Both figures include incidents in which another event may have come first. Of the collisions in that sample, around 53 percent were with stationary objects such as trees, guard rails or mailboxes, and around 43 percent were with other vehicles.

The injury data adds: "The OHV overturned in 68 percent of injuries, and the victim was ejected in 83 percent of injuries." These percentages come from the 2024 special study of non-fatal injuries. The report says those results can be generalized only to non-fatal injuries involving ATVs, ROVs and UTVs with four or more wheels or an unknown number of wheels, an estimated 107,500 such injuries in 2024. The study also found drivers were involved in 73 percent of these injuries, passengers in 23 percent and bystanders in 4 percent.

For someone hurt in a rollover or ejection, the data names the same mechanisms that may be at issue in their own crash. It does not say why any one vehicle overturned or why any one person was ejected. A rollover is not proof of a product defect. A single crash may raise different questions:

• whether someone operated the vehicle carelessly, which is a recreational negligence question;
• whether a trail, property or rental site was unsafe, which is a premises question;
• whether a design, manufacturing flaw or warning was inadequate, which is a product-liability question.

Federal regulations describe a defect broadly. Under 16 C.F.R. § 1115.4, a defect is "a fault, flaw, or irregularity that causes weakness, failure, or inadequacy in form or function." That definition can include design, manufacturing, warnings and instructions. Whether any of these theories fits depends on the supported facts of the individual incident, which is the kind of question an attorney's review of the vehicle, the scene and the records can help sort out.

Which States Had the Highest Reported Death Counts

The report states: "The states with the highest number of reported deaths during the 3-year period were California (166), Pennsylvania (140), Georgia (136), Kentucky (136), and West Virginia (115)." Together, the report says, those five states accounted for 693 fatalities from 642 incidents. The report gives the incident count as 671, or around 26 percent of the 2,685 total.

These are reported deaths, with the same qualifier as the national count. A state's place on the list is not a finding about any product, park or operator there.

For a reader, the state that matters is the one where the crash happened, and its ranking says nothing about a person's legal rights. Limitation periods, comparative fault rules, damages rules and wrongful-death rules differ from state to state. Comparative fault is the rule for how a person's own share of fault affects a claim. Each state's rules come from its own statutes and courts, not from the CPSC's count. The site's California ATV accident page covers that state's rules separately. California's rules apply in California only.

Limitations

Limitations of the Federal Data — What the Numbers Cannot Tell You

The report is careful about its own limits.

• The death figures reflect what "CPSC staff is aware of" as of September 2025. The report calls these data anecdotal and says additional incidents may be recognized later.
• The report describes a lag of generally 2 to 3 years between a death and its report to the CPSC. Staff therefore treats 2023–2025 death data as incomplete and excluded those years. These figures are not real-time.
• The injury figures are NEISS estimates from a probability sample, with a margin of uncertainty. They count only emergency department-treated injuries.
• Anyone hoping to see whether OHV injuries are rising or falling will not find that answer in this edition: because of the 2024 change in NEISS sampling, the report states that linear trend analyses for injury estimates, as presented in prior reports, are not currently available.
• Vehicle labels can be unreliable. Among completed in-depth investigations of fatal incidents involving an ROV or UTV, the original death certificate or fatality report described the vehicle as an ATV around 75 percent of the time. The report says "ATV" is frequently used as a catch-all term, and that the actual distribution of injuries among UTVs and ROVs should be considered unknown and may still be underestimated.
• "Associated with" is not a causal finding.

The data does not address who was at fault, whether a vehicle or component was defective, whether a property condition played a part, or whether anyone is legally responsible. A recall is a separate kind of record, and it is not a finding that a product injured anyone. Brand pages on this site, such as Polaris recall notices and Yamaha recall notices, state what each CPSC notice says in its own words.

If you are trying to understand your own crash, the gap between these national numbers and your family's facts is where the real questions sit: the vehicle, the place, the people involved and the state where it happened. An attorney can review how those facts may relate to the different legal theories, and you can request a case evaluation to start that conversation.