The headline analysis

Risk by experience

A popular idea in general aviation is that pilots are most dangerous in a specific early window of their flying career. It's an intuitive story — but the way it's usually shown is statistically broken. Here's the honest version.

Step 1 — the raw counts

Fatal fixed-wing Part 91 accidents (2008–present) by pilot total flight hours, 3,001 with hours on record:

0–50 hrs 77

50–100 hrs 77

100–150 hrs 102

150–250 hrs 195

250–350 hrs 193

350–500 hrs 204

500–1,000 hrs 509

1,000–2,000 hrs 554

2,000–5,000 hrs 538

5,000+ hrs 552

Step 2 — why this chart lies

Read naively, it looks like experienced pilots are the dangerous ones — they have the most accidents. But this counts accidents, not risk. High-time pilots fly vastly more hours than low-time pilots, so they accumulate more accidents simply by being in the air more. To measure risk you have to divide by exposure.

Risk = accidents ÷ exposure. The right denominators are how many active pilots hold each experience level, and how many hours they fly. Without that, any "experience vs. danger" chart — including the famous one — is just a picture of who flies the most.

Step 3 — what the rate actually is

The aviation-safety community does compute the rate properly at the industry level. The AOPA Air Safety Institute's AOPA Air Safety Institute divides fatal accidents by total estimated flight hours each year. As published by AOPA for non-commercial fixed-wing GA:

2017 0.90

2018 0.88

2019 1.02

2020 0.92

2021 0.86

2022 0.79

2023 0.65

Per 100,000 flight hours. Figures are as published by AOPA ASI and reflect their scope (non-commercial fixed-wing GA) and methodology. Shown here as cited industry context, not recomputed by this site. Total-hours figures from the same report where reported. Source: AOPA Air Safety Institute — Richard G. McSpadden Report (35th edition, 2023 data; formerly the Joseph T. Nall Report).

The takeaway: GA's fatal rate has run roughly 0.8–1.0 per 100,000 hours and fell to a reported 0.65 in 2023 — a record low. But notice this is an industry-wide rate. It still doesn't answer our question, because it isn't broken out by pilot experience.

Step 4 — the experience-band rate (the hard, honest part)

Here's the catch, and it's the whole reason the "danger window" debate never gets settled: no public dataset reports flight hours by pilot experience. The FAA's General Aviation Activity Survey samples aircraft, not pilots — so it tells you how many hours Cessna 172s flew, but not how many hours were flown by pilots with 100 vs. 5,000 hours. The denominator we'd need simply isn't published off the shelf.

That gap is exactly why the popular charts mislead: they show accident counts by experience (which we have) without the matching exposure (which no one publishes). The researchers who tackled it properly had to model the exposure — and when they did, they found the elevated-risk period is real but extends well beyond the folk-wisdom window, not a tidy 50–350 hours (see the 2013 study "The 'killing zone' revisited", Accident Analysis & Prevention, and FAA report DOT/FAA/AM-15/03).

Our next step is to publish a defensible experience-band estimate using those modeled denominators plus FAA active-pilot counts — clearly labeled as an estimate, with its assumptions shown. We'd rather show you an honest range than a precise-looking chart built on a denominator that doesn't exist.