BACKGROUND: Information about risk factors of aviation crashes is crucial for developing effective intervention programs. Previous studies assessing factors associated with crash risk were conducted primarily in general aviation, air taxis, and commuter air carriers. METHODS: A matched case-control design was used to examine the associations of geographic region, basic weather condition, and pilot age with the risk of air carrier (14 CFR Part 121) crash involvement. Cases (N = 373) were air carrier crashes involving aircraft made by Boeing, McDonnell Douglas, and Airbus recorded in the National Transportation Safety Board's aviation crash database during 1983 through 2002, and controls (N = 746) were air carrier incidents involving aircraft of the same three makes selected at random from the Federal Aviation Administration's aviation incident database. Each case was matched with two controls on the calendar year when the index crash occurred. Conditional logistic regression was used for statistical analysis. RESULTS: With adjustment for basic weather condition, pilot age, and total flight time, the risk of air carrier crashes in Alaska was more than three times the risk for other regions ladjusted odds ratio (OR) 3.18, 95% confidence interval (CI) 1.35-7.49]. Instrument meteorological conditions were associated with an increased risk for air carrier crashes involving pilot error (adjusted OR 2.26, 95% CI 1.15-4.44) and a decreased risk for air carrier crashes without pilot error (adjusted OR 0.60, 95% CI 0.37-0.96). Neither pilot age nor total flight time were significantly associated with the risk of air carrier crashes. CONCLUSIONS: The excess risk of air carrier crashes in Alaska and the effect of adverse weather on pilot-error crashes underscore the importance of environmental hazards in flight safety.