Two clusters of teenage suicides occurred in Texas between February 1983 and October 1984. The eight suicides from the first cluster and the six from the second were investigated together in a case-control study using a closedresponse questionnaire administered to parents. Three control teenagers were matched with each case subject on the basis of school district, grade, race, and sex. Questions concerned the direct and indirect exposures of subjects to previous suicides as well as their life circumstances and behaviors. Case subjects were not more likely than control subjects to have had direct exposure to suicide as measured by their acquaintance with a person who committed suicide. Similarly, indirect exposure to suicide through the media was not associated significantly with suicide. Case subjects were more likely to have attempted or threatened suicide previously, to have damaged themselves physically, and to have known someone closely who died violently. Case subjects were more likely to have broken up with their girlfriends or boyfriends recently. They also had moved more often than control subjects, attended more schools, and lived with more parent figures. Identification of high-risk youths through knowledge of relevant risk factors can help to direct preventive services to those young people most susceptible to suicide.