This study examined the life stress and problem‐solving interactional model of suicide proposed by Clum, Patsiokas, and Luscomb (1979). Thirty‐three hospitalized suicidal adolescents were compared with 21 adolescents hospitalized for other psychiatric problems and with 89 controls. The assessment battery was composed of psychological measures, problem‐solving measures, and environmental and family measures. The discriminant analyses revealed that the suicide group could be discriminated from the psychiatric control group but not from the high school control group. Unexpectedly, life stresses did not contribute to the identification of current suicide risk. The results suggest the importance of assessing suicide risk at the time of admission to minimize any subsequent changes in the risk