How many fatal crashes has Waymo had?

What Waymo has published

Waymo publishes periodic safety reports covering the cumulative miles driven autonomously across its commercial service in Phoenix, San Francisco, Los Angeles, Austin, and Atlanta. Through 2026 the company has reported tens of millions of fully autonomous rider-only miles.

Headline claims from Waymo's published safety data:

• No reported rider fatalities during commercial service.
• Substantially lower crash rate than the human-driver baseline for comparable urban environments, particularly for crashes involving injury.
• Lower rates of high-severity crashes specifically.

The methodology is documented; the comparison is against a benchmark derived from human-driver crash data in the same metros where Waymo operates.

What NHTSA's Standing General Order tracks

NHTSA's Standing General Order on Crash Reporting (SGO 2021-01) requires AV operators to report every crash involving their vehicles. This dataset includes:

• Crashes where the Waymo was at-fault.
• Crashes where a human-driven vehicle hit the Waymo.
• Minor contact (parking-lot scrapes) up to severe collisions.

For Waymo specifically, the great majority of SGO-reported incidents involve a human-driven vehicle hitting a Waymo. Often at low speeds in city driving. This is a meaningful editorial distinction: "Waymo was in a crash" is not the same as "Waymo caused a crash."

The Deploy bar on "fatal crashes"

When evaluating fatality claims for any robotaxi service:

1. Who was at fault. Was the AV the primary cause or struck by a human driver?
2. What mode was the vehicle in. Fully autonomous, supervised, or human-driven?
3. Was a rider in the vehicle. Pedestrian or other-vehicle fatalities are different categories than rider fatalities.
4. What baseline are we comparing to. Per-mile, per-trip, per-hour, urban vs highway?

These distinctions matter because the public conversation often collapses them.

Comparison to human-driver baseline

In the US, the human-driver fatality rate is roughly 1.3 per 100 million vehicle miles traveled, with urban driving lower than highway driving. Waymo's reported fatality count of zero across tens of millions of autonomous urban miles is consistent with (or below) that baseline, but the statistical confidence interval at current cumulative mileage is wide.

The fair conclusion: Waymo's published safety record is strong, but the sample size is not yet large enough to make a statistically definitive claim about long-term fatality rates versus human drivers.

Methodological caveats

• Waymo's published comparisons control for metro, road type, and time of day to make like-for-like comparisons against human drivers in the same geographies. This is reasonable but not perfect.
• Self-reporting by the AV operator is the primary data source. NHTSA's SGO is the regulatory check.
• Cumulative miles per metro vary: Phoenix has the longest operational history; newer cities have less data.

Bottom line

As of mid-2026, Waymo has reported zero rider fatalities across its commercial service. The published crash data shows a meaningful safety improvement over human-driver baselines, with the caveat that confidence intervals at current mileage are wider than a single headline number suggests. See what happens if a Waymo gets in an accident and who is at fault if a driverless car crashes for the procedural and liability picture. For methodology canonical references applicable to Waymo safety claims: verified-vs-claimed at within-entity granularity (at-fault vs in-a-crash distinction) + the 9-tier source-quality rubric (NHTSA SGO + Swiss Re actuarial + Waymo published primary-government-record source classification).