A Carrier With 12 Crashes Can Be Safer Than One With Zero. Here's How to Read the Data.

A 400-truck carrier had 12 DOT-reportable crashes in the last 24 months. A 3-truck carrier had zero. A broker screening both carriers flagged the 400-truck fleet and approved the 3-truck operation. Six weeks later, the 3-truck carrier's driver ran a stop sign and caused a fatal accident on a rural highway in Arkansas.

The math the broker missed: 12 crashes across 400 power units over 24 months is a crash rate of 0.03 per truck per year. The national average for carriers of that size is roughly 0.04. That 400-truck carrier was below average. The 3-truck carrier had no crash data at all, which the broker read as "clean" when it actually meant "not enough data to evaluate." Zero crashes is not the same as zero risk. It is the same as zero information.

A carrier's crash history is the record of all DOT-reportable crashes involving that carrier's vehicles over the most recent 24 months, collected by the Federal Motor Carrier Safety Administration (FMCSA) through state-reported data. The raw crash count is publicly available for every carrier. But the count alone is one of the most misread data points in carrier vetting because it lacks two pieces of context that determine whether the number is alarming or expected: the carrier's size and the fault determination for each crash.

Carrier Crash History: How to Read the Numbers

How to Look Up a Carrier's Crash History

Any carrier's crash history is available for free through FMCSA's Safety Measurement System (SMS) website or through the crash history tool, which shows crash records alongside severity breakdowns, trend analysis, and the carrier's inspection data on one screen.

Step-by-Step Lookup

1. Search by DOT number or carrier name on the FMCSA SMS website at ai.fmcsa.dot.gov/SMS or on CarrierBrief's search.
2. Open the Crash Indicator section of the carrier's profile. This shows the total crash count, crash dates, locations, and severity classifications.
3. Note the severity of each crash. FMCSA categorizes every crash as fatal (at least one death), injury (at least one person injured, no fatalities), or tow-away (at least one vehicle towed from the scene, no injuries or fatalities).
4. Check the carrier's power unit count on the same profile. This number is critical for interpreting whether the crash count is high, low, or expected for a fleet of that size.
5. Review the trend direction. Are crashes increasing, decreasing, or stable over the 24-month window? A carrier with 8 crashes in the first 12 months and 2 in the most recent 12 months is on a different trajectory than one with the reverse pattern.

What You Can See vs. What You Cannot

Brokers and shippers can see the raw crash data: dates, states, severity categories, and total counts. What brokers cannot see is the Crash Indicator BASIC percentile. FMCSA restricts this percentile to the carrier and enforcement personnel only. The percentile compares the carrier's crash rate against peers with a similar number of power units, which is the exact context brokers need to interpret the raw number. Without the percentile, you have to do the size-adjustment math yourself.

What Counts as a DOT-Reportable Crash?

A DOT-reportable crash is any traffic incident involving a commercial motor vehicle that results in at least one fatality, at least one injury requiring immediate medical treatment away from the scene, or at least one vehicle towed from the scene due to disabling damage. All three severity thresholds are independent: a crash meeting any one of them is reportable.

The threshold matters because it excludes minor incidents. A fender bender in a truck stop parking lot where both vehicles drive away is not DOT-reportable. A sideswipe on an interstate where the trailer clips a guardrail and a tow truck is called to clear the trailer is reportable, even if nobody was hurt and the damage was cosmetic, because a vehicle was towed.

This creates a specific distortion in the data: tow-away crashes make up the majority of DOT-reportable events, and many tow-aways involve minimal actual severity. A flat tire that disables a vehicle on the shoulder and requires a tow counts the same as a high-speed rear-end collision that totals both vehicles. The crash record does not distinguish between them beyond the severity category.

For brokers evaluating crash data, the severity mix matters more than the count. A carrier with 6 crashes, all tow-away, is in a fundamentally different risk category than a carrier with 6 crashes including 2 fatalities. The raw number is identical. The signal is not.

The Fault Problem: Why Crash Counts Are Inherently Misleading

FMCSA crash records do not include fault determination. Every DOT-reportable crash involving a carrier's vehicle goes on that carrier's record regardless of who caused it. A carrier whose truck was rear-ended at a red light by a distracted passenger vehicle driver receives the same entry as a carrier whose driver caused the accident by running a stop sign.

This is the single most important thing to understand about carrier crash history, and the reason raw crash counts should never be used as a standalone vetting signal.

FMCSA's position, maintained since the CSA program launched in 2010, is that crash involvement carries predictive value for future crashes independent of fault. The agency's statistical research suggests that carriers involved in more crashes, regardless of cause, are more likely to be involved in future crashes. The trucking industry has disputed this methodology for over 15 years, arguing that punishing carriers for crashes they did not cause discourages safe carriers from operating in high-traffic corridors where not-at-fault crashes are statistically inevitable.

The debate is unresolved. What is not debatable: if you look at a crash count without understanding that it includes not-at-fault events, you will misread the data.

How to Account for the Fault Gap

Since fault data is not in the record, brokers need to use indirect signals to assess whether a carrier's crash history reflects their driving quality or their operating environment:

1. Compare crash count to fleet size. A 200-truck carrier with 8 crashes over two years has a crash rate of 0.02 per truck per year. That is below the national average. A 5-truck carrier with 3 crashes has a rate of 0.30. That is ten times higher.
2. Look at the severity mix. A crash history dominated by tow-away events (the lowest severity) in high-traffic states like Texas, California, and Illinois tells a different story than a history with multiple injury or fatal crashes.
3. Cross-reference with inspection data. A carrier with crashes AND high out-of-service rates for vehicle maintenance suggests the crashes may be connected to equipment failures. A carrier with crashes but an impeccable inspection record is more likely experiencing not-at-fault incidents. Read our guide to reading inspection history for how to pull the violation detail.
4. Ask the carrier. For any crash that concerns you, ask the carrier directly what happened. Legitimate carriers can explain the circumstances of their crashes, often with documentation. Carriers who cannot or will not explain their crash history are telling you something.

What Crash Patterns Actually Predict Risk

Not all crash patterns carry equal weight for vetting. Some configurations in the data are strong risk signals. Others are statistical noise. Knowing the difference is the skill that separates experienced vetting from checkbox compliance.

High-Risk Patterns (Act on These)

Multiple crashes in a short window. Three or more crashes within a 6-month period for a carrier with fewer than 50 power units indicates something systemic, whether it is driver quality, equipment condition, or unrealistic dispatch schedules. The probability of a small carrier experiencing multiple crashes in rapid succession through pure bad luck is low.

Fatal or injury crashes with correlated violations. A crash history that includes a fatality or serious injury crash alongside inspection records showing repeated Unsafe Driving violations (speeding, reckless driving, following too closely) is the highest-risk combination in carrier vetting. The crash suggests the outcome. The violations suggest the cause.

Crash concentration in a single corridor or state. If all of a carrier's crashes occurred on the same route or in the same region, it may indicate a driver or group of drivers who operate that lane regularly with risky behavior. Alternatively, it could indicate a genuinely dangerous route segment. Check whether the crashes involved the same driver (you cannot see this in public data, but you can ask the carrier).

Lower-Risk Patterns (Context Required)

Single tow-away crash on a large fleet. A 150-truck carrier with one tow-away crash in 24 months has a crash record that is statistically unremarkable. Most large fleets will experience at least one DOT-reportable event over a two-year window simply due to exposure miles. This is not a signal.

Crashes exclusively in high-incident states. Texas, California, Florida, Georgia, and Illinois account for a disproportionate share of all commercial vehicle crashes nationally because they have the highest truck VMT (vehicle miles traveled). A carrier who operates primarily in these states will have more crash exposure than a carrier running Montana-to-Wyoming routes regardless of driver quality.

Zero crashes on a very small or very new carrier. This is not a positive signal. It is an absence of data. A 2-truck carrier operating for 8 months has not been on the road long enough to generate meaningful crash statistics. Treat zero crashes on small or new carriers the same way you treat "Not Rated" for safety ratings: the field is empty, not clean.

How Long Do Crashes Stay on a Carrier's FMCSA Record?

Crashes remain on a carrier's FMCSA record and affect their Crash Indicator BASIC percentile for 24 months from the crash date. After 24 months, the crash drops off the SMS scoring window entirely and no longer affects any BASIC calculation.

The 24-month window uses the same time-weighting system as all other BASICs:

A fatal crash from last month hits the Crash Indicator roughly three times harder than the same crash would at 18 months old. For carriers who have experienced a serious crash and taken corrective action, the time decay provides measurable improvement within 12 months.

One detail carriers need to know: the crash record itself (the fact that it happened) may remain visible on the carrier's FMCSA profile beyond 24 months as a historical data point. But it stops affecting the BASIC percentile calculation at the 24-month mark. The scoring impact and the record visibility have different retention periods.

How Crash History Affects Insurance and Booking Decisions

Insurance underwriters and brokers both use crash data, but they read it differently and weight it differently in their decisions.

Insurance Impact

Underwriters have access to crash data through FMCSA's public records and through loss-run reports from the carrier's current and prior insurers. A carrier's crash history directly affects their liability insurance premium. The relationship is not linear: the first crash above average for the carrier's size group has a modest impact, but each additional crash compounds the premium increase steeply.

Fatal crashes have the largest impact on insurance pricing regardless of fault. An underwriter pricing a policy for a carrier with a recent fatal crash will factor it in even if the carrier has documentation proving non-fault. The actuarial logic is simple: a carrier involved in a fatal event, for any reason, represents a higher statistical risk than one with no fatal involvement. Fair or not, that is how the math works.

For small carriers, a single serious crash can make liability coverage unaffordable or unavailable. Some underwriters will not write new policies for carriers with any fatal crash in the prior 36 months, regardless of fleet size or fault determination. This is one of the mechanisms through which the no-fault crash data problem translates into real economic harm for carriers.

Broker Vetting Impact

Many brokerages use crash data as a secondary vetting signal after CSA scores and authority verification. The common approach: flag any carrier with crash involvement above a threshold (typically 2+ crashes for carriers under 20 power units, or a crash rate above 0.05 per power unit for larger fleets) for additional review.

The better approach: pair crash data with the carrier's inspection and violation profile. A carrier with 4 crashes and a 15th percentile Unsafe Driving score is a meaningfully different risk than a carrier with 4 crashes and a 72nd percentile. The crashes may be identical. The driving behavior behind them is not.

Using DataQs to Add Context to Crash Records

Carriers who believe their crash records contain errors or lack context can submit a Request for Data Review (RDR) through FMCSA's DataQs system at dataqs.fmcsa.dot.gov. DataQs is FMCSA's formal process for carriers to challenge or annotate data in their safety records.

For crash records specifically, DataQs can accomplish three things:

1. Correct factual errors. If a crash was attributed to the wrong carrier (wrong DOT number), the carrier can request a correction.
2. Add a crash-not-preventable determination. Under FMCSA's Crash Preventability Determination Program (CPDP), carriers can request a review of specific crashes to determine if the crash was "not preventable." If approved, the crash is flagged in the system and excluded from the Crash Indicator BASIC percentile calculation. This does not remove the crash from the record, but it removes its scoring impact.
3. Provide supporting documentation. Police reports, dashcam footage, witness statements, and insurance investigation findings can be submitted to add context that the raw data lacks.

The Crash Preventability Determination Program covers specific crash types: the carrier's vehicle was struck by a wrong-way driver, struck while legally stopped, struck by a vehicle that failed to stop at a traffic control device, or involved in an incident caused by an infrastructure failure. Not all crash types are eligible, but those that are have a high approval rate when documentation supports the claim.

For small carriers with a thin crash record, a single successful CPDP determination can meaningfully change their Crash Indicator percentile. If your one recordable crash in two years was a not-at-fault rear-end collision, the CPDP process can remove its scoring impact entirely.

Frequently Asked Questions

How do I look up a trucking company's crash history?

Search by DOT number on FMCSA's SMS website at ai.fmcsa.dot.gov/SMS and open the Crash Indicator section. It shows crash dates, states, and severity classifications (fatal, injury, tow-away) for the last 24 months. For a faster view with severity breakdowns and trend analysis on one screen, use the crash history tool. All crash data is public and free to access.

Do crashes that aren't my fault show on my FMCSA record?

Yes. All DOT-reportable crashes involving a carrier's vehicle appear on the carrier's FMCSA record regardless of fault determination. FMCSA maintains that crash involvement, independent of cause, carries predictive value. Carriers can mitigate the impact by submitting a Crash Preventability Determination Program (CPDP) request through DataQs for eligible crash types. If approved, the crash remains on the record but is excluded from the Crash Indicator BASIC percentile calculation.

How long do crashes stay on a carrier's record?

Crashes affect the Crash Indicator BASIC percentile for 24 months from the crash date, with time-weighted scoring: full weight for months 0 to 6, two-thirds weight for 6 to 12 months, and one-third weight for 12 to 24 months. After 24 months, the crash drops from the BASIC calculation entirely. The crash event itself may remain visible as a historical record beyond 24 months but no longer affects any safety score.

What is the Crash Indicator BASIC?

The Crash Indicator is one of seven BASIC categories in FMCSA's Safety Measurement System. It measures the frequency and severity of a carrier's DOT-reportable crashes over 24 months, compared to similar-sized peers. The Crash Indicator percentile is the only BASIC that is not publicly visible. Only the carrier and FMCSA enforcement personnel can view it. Brokers can see the raw crash data but not the percentile ranking. The intervention threshold is 65%.

Can a broker be liable for booking a carrier with crash history?

Yes, through negligent selection claims. If a carrier's crash records showed elevated risk and a broker failed to review or act on that publicly available data before tendering a load, the broker faces liability exposure if the carrier subsequently causes an accident. Several court decisions have found brokers partially liable in exactly this scenario. Checking crash history alongside CSA scores and authority status is the floor of reasonable due diligence.

How many crashes is too many for a carrier?

That depends entirely on fleet size. A 300-truck carrier with 10 tow-away crashes in two years has a crash rate of approximately 0.017 per power unit per year, which is below the national average. A 5-truck carrier with 3 crashes has a rate of 0.30, which is roughly 18 times higher. Evaluate crashes as a rate per power unit, not as a raw count. Also examine the severity mix: 10 tow-aways are a different risk profile than 2 fatalities and 8 tow-aways.

Why does FMCSA count crashes regardless of fault?

FMCSA's statistical research indicates that crash involvement, independent of fault, correlates with future crash risk. The agency's position is that carriers with higher crash exposure are more likely to experience future crashes regardless of who caused prior events. The trucking industry has contested this methodology since CSA launched in 2010. The Crash Preventability Determination Program was introduced as a partial concession, allowing carriers to exclude specific not-preventable crashes from their BASIC scores, but the underlying all-crashes-count methodology remains unchanged.

What is the Crash Preventability Determination Program?

The Crash Preventability Determination Program (CPDP) is an FMCSA program that allows carriers to request a review of specific crash types to determine whether the crash was "not preventable." Eligible crash types include being struck by a wrong-way driver, being struck while legally stopped, and crashes caused by the other vehicle's failure to obey traffic signals. If FMCSA determines the crash was not preventable, it remains on the carrier's record but is excluded from the Crash Indicator BASIC percentile calculation. Requests are submitted through the DataQs system at dataqs.fmcsa.dot.gov.

The 400-Truck Fleet and the 3-Truck Carrier

That broker who flagged the 400-truck carrier for 12 crashes and approved the 3-truck operation with zero made the exact decision the raw data encouraged. Twelve is a bigger number than zero. The record looked worse. The math said otherwise.

Twelve crashes across 400 trucks over two years is a rate below the national average. Zero crashes across 3 trucks over 8 months is not a clean record. It is 8 months of insufficient data on an operation too small and too new to have generated meaningful crash statistics. The broker screened out the carrier with a demonstrated safety record and booked the one with no record at all. Six weeks later, a family in Arkansas paid for that decision.

Pull the crash history. Then pull the power unit count. Divide. Compare to the fleet size average. Read the severity mix. Cross-reference with inspection data. And when the crash count is zero on a small carrier, treat it as the absence of evidence it is, not the evidence of absence you want it to be.