OBJECTIVE: The authors previously identified suspected risk factors for traumatic events related to posttraumatic stress disorder (PTSD) on the basis of data gathered retrospectively. In this study, they tested that model prospectively. METHOD: A random sample of 1,200 persons was drawn from all 21- to 30-year-old members of a large health maintenance organization. In 1989, 1,007 of these persons were interviewed, and suspected risk factors were measured. In 1992, 979 were reinterviewed, and the 3-year incidence of exposure to traumatic events was ascertained. RESULTS: Nineteen percent of the sample reported traumatic events during the 3-year follow-up. A history of past exposure to traumatic events signaled an increase in the liability to exposure during follow-up, independent of suspected risk factors. Two predictors of exposure, neuroticism and extroversion, identified retrospectively, also predicted exposure prospectively. The odds for exposure among males and persons with less than a college education were marginally significant. Early misconduct and a family history of psychiatric disorder-predictors of exposure in the retrospective data-were not significant predictors at 3-year follow-up. Blacks had a higher incidence of exposure during follow-up than whites. An exploratory reanalysis suggested that the discrepancy between the retrospective and prospective results may be explained by the inclusion of childhood exposure in the lifetime retrospective inquiry. CONCLUSIONS: The assumption that PTSD-related traumatic events are random phenomena was unsupported. Among young adults, those with less education, blacks, and those with high neuroticism and extroversion scores are more likely than others to be exposed to traumatic events and are thus at greater risk for PTSD.