Drug courts have always used a primitive form of screening; however, it has consisted of establishing little more than a substance-abuse problem, a desire for help, and a lack of violent criminal history. Screening, however, should not be used merely as a tool for excluding offenders deemed to be at high risk for failure. Screening should be a form of assessment, which could help identify drug court participants in need of specialized assistance or alternative programs. Accurate screening could increase efficiency in drug treatment. The current study drew from South Carolina county-level drug court records to examine success rates for participants with particular background characteristics. Risk factors for drug court participants were identified by using chi-square and discriminant function analysis. Recidivism, crack as the drug of choice, existence of criminality prior to drug use, and prior drug treatment were significantly related to drug court failure. Although this study had a number of methodological flaws, the findings suggest that a well-designed screening mechanism could assist a drug court administrator in increasing success rates. 2 tables, 6 notes, and 33 references