Your car kills people without ever crashing

Ten Americans died with Tesla's FSD engaged in ten months. Vehicle exhaust kills roughly 385,000 people a year worldwide — and not one of them counts as a car accident, because nobody hit anything.

Here is a number nobody put on a chyron.

A global study of on-road vehicle exhaust attributed roughly 361,000 premature deaths in 2010 and about 385,000 in 2015 to tailpipe emissions — fine particulate matter and ozone produced by cars, vans, buses and trucks burning liquid fuel. Around 70% of that burden landed in just four markets: China, India, the European Union and the United States. Diesel alone accounted for nearly half of it, about 181,000 deaths.

Now hold that against the number this entire website exists to put in front of you: in the ten months from January to October 2025, ten people died on American roads with Tesla's Autopilot or Full Self-Driving verified as engaged.

Ten. Against three hundred and eighty-five thousand.

The deaths that don't count as accidents

The reason you have never seen the 385,000 figure in a headline is structural, and it is worth understanding because it explains most of what is wrong with how you have been informed.

A crash is an event. It has a time, a place, a photograph, a mangled object, a name, and a grieving family who will talk to a reporter. It is, in the trade, a story.

Tailpipe mortality is a distribution. It is a slight upward shift in the probability that a sixty-eight-year-old man in Bakersfield has the heart attack this winter rather than in four years' time. Nobody photographs it. Nobody attends the scene. The death certificate says cardiac arrest, not Chevrolet. There is no wreckage and there is no villain, so there is no coverage — and because there is no coverage, there is no fear.

This is not a conspiracy. It is worse than a conspiracy, because a conspiracy could be exposed and stopped. It is an incentive structure, and it runs on its own.

What is actually in the air

The mechanism is not mysterious. Combustion engines emit fine particulate matter — PM2.5 — small enough to cross from your lungs into your bloodstream. Transportation is responsible for up to roughly a quarter of US premature deaths attributable to PM2.5.

And it does not stop at your lungs. There is strong evidence that breathing PM2.5 impairs neurological function and raises the risk of Alzheimer's disease, and that it affects the developing fetus. Nearly half of Americans live in counties with unhealthy levels of air pollution, with cars and trucks a leading cause.

Sit with that for a second, because it is the part people skate past. The exhaust is not merely killing pensioners with bad hearts a few years early. It is a plausible contributor to dementia and to how children's brains develop before they are born.

That is what is coming out of the thing on your driveway. Every day. While you are frightened of a battery.

"But the long tailpipe"

Here comes the reply, right on cue, from a man who has never opened a lifecycle analysis in his life: the electricity comes from somewhere, you've just moved the tailpipe.

It is a fair question that was answered years ago and keeps being asked as though it hasn't been.

Yes — building an EV emits more carbon than building a petrol car, because of the battery. That is real and nobody serious denies it. The question is how long it takes to pay that back, and the answer is now embarrassingly short:

• In the United States, on average grid mix, the breakeven lands somewhere around 13,500 to 21,300 miles — for most drivers, well under two years.
• In Europe the figure is roughly 17,000 km, about one to two years of normal driving.
• Over a full 250,000 km life, a medium battery-electric car built today produces 27–71% lower lifecycle CO2 than the equivalent combustion car.
• The EPA states plainly that EVs typically have a smaller carbon footprint than gasoline cars, even accounting for electricity generation.

And here is the part that ends the argument: the grid gets cleaner every year, so an EV's lifetime emissions fall while you own it. A petrol car's emissions are fixed on the day it leaves the factory. It will never burn cleaner than it does today. Not once. Not ever.

The long-tailpipe argument was never a finding. It was a hope — a hope that the sums would come out the other way — and the sums came in, and they didn't, and the argument survived anyway because it was never about the sums.

Why the argument survives

People who repeat the long-tailpipe line are not, in the main, lying. They heard it somewhere credible-sounding, it flattered a decision they had already made about their own car, and they never went back to check. That is not stupidity. That is ordinary human behaviour and every one of us does it about something.

But there is a difference between not knowing a thing and not wanting to know it, and the tell is what happens when you are shown the numbers. If you read the paragraph above and your first instinct was to look for a reason the studies must be wrong rather than to check whether they are — that instinct is worth examining. It is the same instinct that kept lead in petrol for fifty years after the science was in.

The scoreboard

So let us put it plainly, since nobody else will.

The technology you have been trained to fear killed ten Americans in ten months with its most-maligned feature engaged.

The technology you consider normal, safe, familiar and beyond question kills on the order of 385,000 people a year worldwide — silently, without a crash, without a headline, without a single reporter ever standing in front of a camera and saying the manufacturer's name.

One of those is a story. The other is a body count.

You have been sold the wrong one, and you have been buying it enthusiastically.

The next time somebody tells you they don't trust electric cars because of the batteries, ask them how many people they think tailpipes killed last year. Then watch what happens when you tell them.

Straight answers

How many people do vehicle tailpipe emissions kill each year?

A global study of on-road vehicle exhaust attributed approximately 361,000 premature deaths in 2010 and approximately 385,000 in 2015 to tailpipe emissions of fine particulate matter (PM2.5) and ozone. Around 70% of that burden occurred in China, India, the European Union and the United States. Diesel exhaust accounted for nearly half, roughly 181,000 deaths.

How many people died with Tesla FSD or Autopilot engaged?

In the United States between 1 January and 31 October 2025, ten deaths occurred in crashes where Tesla Autopilot or Full Self-Driving was verified as engaged, according to the tesladeaths.com dataset, which cross-references NHTSA's Standing General Order database, press reports and court records. Over the same window roughly 30,411 people were killed on American roads in vehicles that were not Teslas.

Is the 'long tailpipe' argument against EVs correct?

No. Building an electric car does emit more carbon than building a petrol car because of the battery, but the payback period is short. In the United States on average grid mix the breakeven is roughly 13,500 to 21,300 miles, typically under two years of driving. Across a 250,000 km lifetime a medium battery-electric vehicle produces about 27 to 71 percent lower lifecycle CO2 than an equivalent combustion vehicle. The grid also gets cleaner over time, so an EV's lifetime emissions fall while it is owned, whereas a combustion car's are fixed at manufacture.

Does breathing vehicle exhaust cause anything other than lung problems?

Yes. Beyond respiratory and cardiovascular harm, there is strong evidence that exposure to PM2.5 impairs neurological function and increases the risk of Alzheimer's disease, and that it affects the developing fetus. Transportation is responsible for up to roughly a quarter of US premature deaths attributable to PM2.5.