The likelihood that a given drug will be misused is related to its ability to alter mood, feeling, thinking, and perception in a manner that is liked by substance abusers. The subjective and behavioral effects of diazepam (10, 20, and 40 mg), pentobarbital (120 and 240 mg as a positive control), and placebo (negative control), were evaluated in 12 subjects with histories of substance abuse by use of a double-blind, Latin square crossover study design. Drug administration was separated by a minimum of 3 days. Pharmacodynamic measures included subjective (euphoria, subject liking, sedation, and symptoms) and behavioral (signs and observed liking) responses. The time course and profile of subjective and behavioral responses were similar for diazepam and pentobarbital. Valid relative potency estimates for the pharmacodynamic measures indicated that diazepam is approximately 10 times as potent as pentobarbital. The study indicates that the reinforcing effects of diazepam are similar to pentobarbital in substance abusers.