One in Three Dead Motorcyclists Weren’t Licensed to Ride. The System Doesn’t Care.

To legally ride a motorcycle in most American states, you take a written test, pass a skills evaluation on a closed course, and pay somewhere between $15 and $30. In some states you can skip the riding test entirely by completing a weekend safety course. It is, by design, one of the lowest bars in American transportation regulation.

In 2023, 35% of motorcyclists killed on American roads hadn't cleared it.

That's 2,094 riders out of 6,025 fatally injured motorcycle operators, according to FARS data compiled by IIHS. No valid motorcycle license, no endorsement, nothing. Dead on a machine they weren't authorized to operate.

For car drivers, the equivalent number is 18%. Bad enough. But 18% looks like a functioning system next to 35%.

COVID Broke It and Nobody Fixed It

From 2014 through 2019, the unlicensed fatality rate among motorcycle riders held remarkably steady at 30 to 31%. Not good, but stable. About 1,500 unlicensed riders died per year, year after year, like clockwork.

Then 2020 hit. DMVs closed. Riding tests got backlogged. And a lot of people who were stuck at home bought motorcycles. Registrations surged. Sales of sub-500cc bikes (the beginner segment) jumped double digits.

In 2020, the unlicensed fatality share spiked to 36%. By 2021, it was 38%. More than 2,000 unlicensed riders dead in a single year for the first time.

By 2023, it had only settled back to 35%. Not 30%. Not 31%. The pre-COVID baseline never came back. We added roughly 500 extra unlicensed motorcycle deaths per year, permanently, and moved on.

A 6,000% Increase Nobody Talks About

Set the licensing problem aside for a second. Look at who's dying.

In 1975, five motorcyclists aged 70 or older died on American roads. Five total, nationwide. In 2023, that number was 295. A 5,800% increase across five decades, and the curve is still climbing.

IIHS Fatality Facts breaks out the age trajectory clearly. In 1982, 4% of dead motorcyclists were 50 or older. By 2023, roughly a third were. Meanwhile, the Motorcycle Industry Council reports that the median motorcycle owner age climbed from 27 in 1980 to 50 in 2018. The rider population aged. The fatality demographics followed.

Among people 70 and older, motorcycle deaths as a share of all motor vehicle deaths went from under 1% in 1975 to 5% in 2023. Older Americans are riding more and dying more on motorcycles, at a rate that dwarfs the growth in any other vehicle category for that age group.

Cross these two trends. You have an aging rider population on the deadliest common vehicle per mile traveled (28 times the fatality rate of cars, per NHTSA 2025 publication 813732), and a licensing system that fails to screen out more than a third of the people who die.

What "Unlicensed" Actually Means

Fair objection: "unlicensed" isn't the same as "unskilled." Plenty of experienced riders let endorsements lapse. Some ride on learner permits. Some have a valid car license but never bothered with the motorcycle endorsement because their state doesn't aggressively enforce it.

This is true. And it doesn't help the argument as much as it sounds like it should.

NHTSA research on motorcycle crash characteristics shows that unlicensed operators are significantly overrepresented in single-vehicle crashes. That's the crash type where the rider runs off the road, fails to negotiate a curve, or loses control on gravel. No other vehicle involved. Just the rider and the road and a mistake.

Unlicensed operators are also significantly more likely to be alcohol-impaired at the time of the fatal crash. This isn't a population that overwhelmingly consists of experienced riders who forgot to renew a form. It's a population where risk-taking behavior clusters. Skip the license, skip the helmet, skip the sobriety.

That clustering matters because it suggests the licensing system isn't just failing as a screening mechanism. It's failing as a signal. In states with weak enforcement, not having a motorcycle license carries roughly the same practical consequence as not having a fishing license. The risk calculus for a rider is: will I actually get pulled over and checked? In most places, no.

17 States Have Figured Out Helmets. Zero Have Figured Out Licensing.

Seventeen states and D.C. require all motorcycle riders to wear helmets. Helmets reduce death risk by 37 to 42%, per IIHS. In an October 2024 analysis, IIHS estimated that more than 20,000 motorcyclists would have survived since the mid-1970s if every state had a universal helmet law.

Helmet laws are the closest thing to a proven policy intervention for motorcycle deaths. They work. States that have them see lower fatality rates. States that repeal them see immediate spikes.

Licensing gets none of this attention. Nobody runs the counterfactual on what would happen if every state actually enforced motorcycle endorsement requirements. Nobody tracks licensing compliance the way they track helmet use rates. NHTSA publishes helmet use surveys annually. There is no equivalent survey for license compliance among active riders.

We don't even know the denominator. How many unlicensed riders are on the road right now, alive, not yet in the FARS database? Nobody counts them. We only count them after they die.

Women Own 19% of Motorcycles. They're 4% of the Dead.

One more number that contextualizes the licensing gap.

According to the Motorcycle Industry Council, women represent 19% of motorcycle owners. But in 2023, women were only 4% of motorcycle operators killed in crashes. Ninety-one percent of motorcycle passengers killed were women, which is its own grim statistic, but as operators, women die at roughly one-fifth the rate their ownership share would predict.

The simplest explanation is behavioral, not mechanical. Women riders are less likely to speed, less likely to ride impaired, and less likely to ride without a valid license. If you want to know what a population of motorcycle riders who actually follow the rules looks like in fatality data, women are the closest proxy we have. And they die at a fraction of the rate.

This doesn't prove that licensing is the variable. It suggests that the behaviors that correlate with being unlicensed (speeding, impairment, single-vehicle crashes) are doing most of the killing.

6,335 and Counting

Total motorcycle fatalities in 2023: 6,335. Highest number ever recorded in the United States. Not highest rate per registered vehicle or per mile traveled. Highest absolute count. More riders died in 2023 than in any previous year on record.

Car occupant deaths have been trending down per mile driven for decades. Airbags, crumple zones, electronic stability control, automatic emergency braking. Cars got safer. Motorcycles didn't. A 2026 model-year Harley-Davidson offers roughly the same crash protection as a 1976 model-year Harley-Davidson: your body and whatever gear you chose to wear.

Into that protection gap, add a licensing system with a 35% failure rate among the people it's supposed to protect, an aging rider demographic that's pushing fatality counts in the 70+ age group to levels nobody anticipated, and 33 states that still don't require adults to wear helmets.

We have a regulatory mechanism that's supposed to keep unprepared riders alive. A third of the dead weren't in the system. The number jumped 40% during COVID. It never came back down. And nobody is working on it because motorcycle licensing isn't the kind of policy fight that generates headlines or political capital.

It's a fee. It was always a fee.

Methodology note: Fatality data from NHTSA's Fatality Analysis Reporting System (FARS), 2023, as compiled by IIHS Fatality Facts: Motorcycles and ATVs (published July 2025). License status reflects FARS coding of motorcycle operator license validity at time of fatal crash. "Unlicensed" includes no license, suspended/revoked license, and no motorcycle endorsement on an otherwise valid license. Age distribution data from IIHS Fatality Facts: Older People (2023). Demographic and ownership data from Motorcycle Industry Council Owner Survey (2019). Per-mile fatality risk ratio (28x) from NHTSA Traffic Safety Facts publication 813732 (2025). Helmet effectiveness and universal helmet law analysis from IIHS Topics: Motorcycles and IIHS press release, October 2024. FARS captures only fatal crashes; non-fatal motorcycle crash data is less consistently reported across states. License compliance rates among living riders are not systematically measured, limiting the ability to calculate unlicensed rider fatality risk relative to exposure. The COVID-era spike in unlicensed deaths may reflect both DMV disruptions and changes in rider behavior/demographics; these effects cannot be fully disentangled with available data.