The present study was a cross-sectional examination of the causes of adolescent participation in risky behaviors, using a diverse sample of adolescents from a Midwestern state. This research compared the relative effects of a traditional explanation of deviant behavior, differential association theory, with a modern derivative, nonsocial reinforcement theory. Quantitative data were collected via questionnaires in the spring of 1998 from 86 students at a public high school, 12 students from a for-profit high school, and 16 students from a detention facility for status offenders. Respondents were asked to indicate the extent of their involvement with cigarettes, alcohol, and various drugs. The multiple linear regression results indicated that while both differential association and nonsocial reinforcement had statistically significant effects on involvement with drugs and alcohol in this sample, nonsocial reinforcement theory had a stronger explanatory effect than differential association. Further, contrary to official statistics, neither gender, race, or social class had a statistically significant effect upon engagement in the risk-taking behaviors in this study. (Contains 25 references.) (Author/KB)