Examined the relationship between 26 dichotomous risk factors and drug use in derivation (N = 1,352) and cross-validation (N = 1,309) samples of seventh graders in the public school system of a large southeastern city. The majority of students was African American, many came from low-income, single parent families. A total of 20 risk factors representing a variety of variables was significantly related to at least one category of drug use in both samples. Regression analyses identified a subset of 11 risk factors with minimum overlap. The simple sum of these 11 risk factors was significantly associated with prevalence of use for cigarettes, beer and wine, hard liquor, marijuana, and other drugs. The total number of risk factors also showed a curvilinear relationship with the frequency of 30-day use for each category of drug. Theoretical and practical implications of these findings are discussed.