This study investigated whether previously reported links between sleep and risk taking among adolescents (E. M. O'Brien & J. A. Mindell, 2005) are associated--concurrently, longitudinally, or both--with sleep or underlying depression. The present study analyzed data from a nationally representative sample of 4,353 adolescents in the United States who had participated in Waves I and II of the National Longitudinal Study of Adolescent Health (J. R. Udry, 1998). In the present study, grade, gender, depressive symptoms, and possible insomnia served as predictor variables to calculate the odds ratios for 4 categories of risky behaviors. After the authors accounted for other predictors, possible insomnia significantly predicted smoking, delinquency (within each wave but not longitudinally), and drinking and driving (Wave II and longitudinally). Possible insomnia may contribute to some types of risky behaviors over the effects of depressive symptoms. (Contains 5 tables.)