Zero Crashes Isn't a Green Light — It Might Be a Blind Spot

A broker I know used to describe her carrier vetting process like this: "I pull the SAFER. If it says zero crashes, I move on. If it has crashes, I read them. Straightforward."

She's good at her job. She thinks carefully about what she books. And she got sued on a $78,000 load where the carrier had zero SAFER crashes, 18 months of active authority, and three BASIC scores showing no flags at all.

The driver had four employers in twelve months. The carrier ran a pre-employment query on the FMCSA Drug and Alcohol Clearinghouse — clean. They didn't run a full PSP report. Didn't request the driver's complete CDL history. His commercial license had been suspended 90 days, eighteen months earlier, for a DUI-adjacent violation in a personal vehicle — the kind that shows on a PSP but often doesn't trigger the Clearinghouse. Zero crashes in SAFER. One fatal crash on load number three of this working relationship.

None of this means her vetting was negligent by the legal standard. That's what her attorney is now working on. What it means is that "zero crashes" was one data point, not a conclusion. She treated it like a conclusion.

What FMCSA Counts as a Crash

Start here because it's not what most people assume.

Under 49 CFR § 390.5, an "accident" that FMCSA collects and tracks requires at least one of three outcomes: a fatality; bodily injury to a person who receives medical treatment away from the scene; or one or more motor vehicles incurring disabling damage requiring a tow. That's it. That's the legal threshold.

Everything below it is invisible to the FMCSA crash database. It doesn't factor into the Crash Indicator BASIC. It doesn't show up when you pull a SAFER snapshot. A driver who rear-ends a car at 15 mph and the other vehicle drives away — not DOT-reportable. A driver who backs into a loading dock door, dents it, no vehicle disabled — not reportable. A driver who sideswipes a guardrail, scrapes the trailer, keeps rolling — not reportable. A multi-vehicle incident where everyone refuses EMS at the scene and everybody drives home — potentially not reportable, depending on state reporting rules.

None of those show up in SAFER. All of them potentially show up in the carrier's § 390.15 accident register, which the carrier is required to maintain internally but which is not publicly accessible the way SAFER is.

So the SAFER crash register doesn't tell you how many times a carrier's drivers have had incidents. It tells you how many times those incidents crossed a specific legal threshold. A carrier running tight lanes, heavy freight, and a mediocre safety culture can go years without a reportable crash — just because their incidents stay below the tow-away threshold.

Why Zero Is Actually Harder to Read Than Three

A carrier with three crashes in 24 months gives you information. You can look at the severity breakdown — fatal, injury, tow-away only. You can check whether any were flagged as preventable. You can cross-reference the Crash Indicator BASIC against their Vehicle Maintenance score and see if there's a pattern.

A carrier with zero crashes gives you an absence. Absence is harder to read.

There are four distinct reasons a carrier might show zero FMCSA-reportable crashes:

First, they're genuinely safe. Their drivers respect hours-of-service rules, their equipment passes inspections, their safety manager runs a real program. The absence of crashes reflects actual operational discipline.

Second, they're new. A carrier with eight months of authority — call them MC-2847193, DOT-3981024 — hasn't had enough load-days for a statistically expected crash to materialize at any underlying risk level. Zero at that sample size proves nothing. It's what you'd expect.

Third, they run light loads on short routes. A carrier doing 20-mile regional shuttles between two facilities has vastly lower crash exposure than an OTR carrier running 140,000 miles a year. The same zero-crash record on both carriers doesn't carry the same meaning.

Fourth, their incidents stay below the reportable threshold. Backing accidents, dock scrapes, minor contact — none of which required a tow or sent anyone to the hospital. The carrier's § 390.15 accident register has entries the SAFER snapshot doesn't show.

Before Montgomery v. Caribe Transport II, LLC — the unanimous Supreme Court decision from May 14, 2026, which held that the FAAAA does not preempt state-law negligent-selection claims against brokers — this distinction mattered mostly to the carrier's own safety management. Post-Montgomery, it's part of your due-diligence problem. The standard for a broker now is ordinary care. Did you take reasonable steps to understand what "zero crashes" actually meant for this specific carrier? Or did you accept the absence of data as positive evidence of safety?

The Inspection Frequency Cross-Check

The fastest proxy for evaluating a zero-crash carrier is inspection count.

Pull the SAFER snapshot and look at the Inspections section. How many inspections does this carrier have in the past 24 months?

If they have 36 months of active authority and four total inspections, you have a thin file. Their zero-crash record exists in a data vacuum. Four inspections across three years of operations is not a population — it's a rounding error. A carrier who hasn't been looked at can't fail an inspection they haven't had.

A carrier with 36 months of authority and 160 inspections is different. Five or six inspections a month suggests they're running high-mileage interstate lanes and getting pulled in at weigh stations regularly. Zero crashes across 160 inspection events, with enforcement attention that frequent, tells you something real. The sample is large enough to be informative.

This is exactly the same logic that makes OOS rates at low inspection counts unreliable. A 14% vehicle OOS rate built on seven inspections doesn't mean the same thing as a 14% OOS rate built on 110 inspections. Same percentage, completely different confidence level. Zero crashes works the same way. The denominator — how much exposure and enforcement observation exists behind the zero — determines whether the zero means anything.

Pair the crash count with the inspection count. If both are zero or near-zero, you have a carrier the FMCSA system simply hasn't seen much of. Combine that with short authority age and you have a carrier who is, by definition, mostly unknown. That's not a green light. That's a carrier whose file is thin by the nature of their operating history.

The T-Call Question That Matters Here

Under 49 CFR § 390.15(b), every motor carrier must maintain an accident register for at least three years. The register has to include each DOT-reportable accident: date, location, driver name, number of injuries, number of fatalities, whether hazmat was involved.

You don't have automatic access to that register as a broker. You're not an enforcement official. But you can ask about it — and asking, on the record, changes your documentation in a specific way.

On the T-call for any carrier with a zero or thin crash history, I ask one direct question: "How long has it been since your last DOT-reportable crash?"

The answers can go several ways.

A carrier who says "we've never had a reportable crash" with 36 months of authority and 140+ inspections — I note that in the file and it's meaningful.

A carrier who says "we've never had one" and they're nine months old — that's expected. The question shifts to their inspection record.

A carrier who pauses, hedges, or gives something vague — I ask again. "Has anything happened in the last six months I should be aware of?" If dispatch is uncomfortable with a direct question about their crash history, that's itself a signal.

Whatever the answer is, it goes in my T-call notes. Under the negligent-selection standard that Montgomery confirmed, what matters is whether you took reasonable steps to inquire. A broker who pulled SAFER, saw zero crashes, and wrote "clean" relied on an absence of data. A broker who pulled SAFER, noted the inspection count was low, asked the direct T-call question, and documented the response applied actual judgment. Those are different positions in a deposition.

New Authority Is a Different Conversation

If you're vetting a carrier under 18 months of authority, zero crashes is essentially meaningless in either direction. It's what you'd expect. You simply haven't been operating long enough for the number to carry weight.

What matters on new authority is the inspection record they DO have. What violations showed up? Were any out-of-service? What specific regulations were cited? A brand-new carrier with eight inspections and 31% vehicle OOS is telling you something real about their equipment. A brand-new carrier with four inspections and zero violations has passed a small test but you don't yet know what full operations look like.

New entrants go into FMCSA's safety assurance program under 49 CFR Part 385 Subpart D — which includes a mandatory safety audit in the first year and targeted inspection exposure at weigh stations. The enforcement attention during that period is not random. New carriers get scrutinized more. An 11-month-old carrier with a clean inspection record has cleared a higher bar per inspection than a five-year carrier whose inspections are routine.

That context is worth noting. It doesn't mean you book every new carrier with clean inspections. It means you interpret their record with the enforcement context in mind.

How I Document This

When I approve a carrier with zero crashes, the note in my file reflects the work I did to evaluate what that zero actually means — not just that I saw it.

Something like: "Carrier shows 0 DOT-reportable crashes in 24-month SAFER window. Authority 31 months. 44 inspections in 24 months. Vehicle OOS rate 9%. T-call [date]: dispatcher confirmed no DOT-reportable crash in the past 12 months. Zero crashes interpreted as consistent with inspection frequency and operating history. Approved."

That note takes 45 seconds to write. It shows that I understood zero crashes is a data point requiring context — not a conclusion. It documents the inspection cross-check, the T-call question, and the response. It doesn't prove the carrier is safe. What it proves is that I applied judgment rather than assumption to an absence of data.

That's ordinary care. That's what the file needs to reflect.

The brokers who get into trouble after Montgomery aren't the ones who approved a dangerous carrier despite obvious flags. They're the ones who didn't notice they had no information — and mistook the absence of bad data for the presence of good data. Zero crashes can be exactly that: not a clean record, but an unread one.

— Mason Lavallet

Founder, DOTScreener.com