Waymo Publishes Safety Data Showing 82% Fewer Injury Crashes Than Human Benchmark

Waymo’s latest safety update puts a large claim behind its driverless rides: across 271.3 million miles through June 2026, its vehicles had 82% fewer injury-causing crashes than a human-driver benchmark for the same distance in its operating areas. The company calculates that gap as 841 crashes. It is a benchmark comparison, not a count of crashes Waymo directly observed and prevented.

How the comparison works

The mileage covers rides without a human driver in Los Angeles, the San Francisco Bay Area, Phoenix, Austin and Atlanta. Waymo compares its crashes with the number expected if drivers at the local human benchmark rate traveled the same distance on surface streets. Its counts include crashes regardless of who was at fault.

That distinction is essential to the 841 figure. It is the calculated difference between Waymo’s recorded crash total and the human benchmark’s expected total over those miles. It does not identify 841 particular collisions that a Waymo vehicle avoided. Waymo says its human benchmarks are adjusted for differences in the mix of driving, since streets within a city do not all pose the same challenge.

A wider set of crash measures

The injury figure is not Waymo’s only measure. It reports a larger percentage reduction for crashes causing serious injury or worse, alongside a reduction in crashes that deployed an airbag in any vehicle. These categories describe different outcomes; the figures should not be added together as though they were separate sets of crashes.

82% fewerCrashes causing injury

Waymo calculates 841 fewer injury-causing crashes than the human benchmark over the same distance.

95% fewerSerious injury or worse

Waymo calculates a difference of 55 serious-injury-or-worse crashes.

82% fewerAirbag deployment

Waymo calculates 358 fewer crashes involving an airbag deployment in any vehicle.

The dashboard also breaks out crashes involving people outside cars. Against its human benchmarks, Waymo reports 93% fewer pedestrian crashes with injuries, 86% fewer cyclist crashes with injuries and 82% fewer motorcycle crashes with injuries. Those comparisons extend the question beyond passenger safety to others sharing the street, while remaining company-reported results.

One total, different city exposures

The five areas contribute very different amounts of driving to the combined result. Phoenix accounts for 92.121 million rider-only miles and the San Francisco Bay Area for 82.421 million; Atlanta accounts for 8.624 million. A reader looking at a city’s result therefore needs to consider its mileage and local comparison, rather than treating the five areas as equal slices of the overall figure.

For serious-injury-or-worse crashes, Waymo’s displayed all-area rates are 0.01 incidents per million miles against 0.21 for the human benchmark. But the city-level charts also show uncertainty ranges, called 95% confidence intervals. Waymo displays city comparisons only where it says it has enough miles for statistical comparison. A point estimate of zero in a city is not a promise that its future crash rate will remain zero.

What the update can answer

Waymo presents the dashboard as a retrospective check on driving already completed, not its sole test of whether its system is ready to operate. It says readiness is assessed through a separate safety case and that the dashboard is updated in line with federal crash-reporting timelines. The new mileage strengthens the view of performance in these operating areas. Whether a local human-driver benchmark is enough to settle the broader safety question remains a judgment the headline percentages cannot make for readers.