Rhesus monkeys were trained in a discrete-trials choice procedure and allowed to choose between food delivery (1-16 pellets; 1 g/pellet) and intravenous injections of cocaine (0.03-0.56 mg/kg/injection; N = 4) or procaine (1.0-10 mg/kg/injection; N = 4) during daily 3-h sessions. Injections were available as the alternative to food. When the amount of food available as the alternative to drug was held constant and dose of drug was varied, the frequency of drug choice and total drug intake increased in a dose-related fashion for both cocaine and procaine. For both drugs, when the amount of food available as the alternative to drug was increased and the dose of the drug was held constant, the frequency of drug choice and total drug intake decreased. Thus, increases in the magnitude of an alternative non-drug reinforcer decreased cocaine and procaine self-administration. Further, the results suggest that while increasing the magnitude of the alternative reinforcer decreased the potency of cocaine as a positive reinforcer, the reinforcing efficacy of procaine was decreased. Because drug use by humans typically occurs in a context in which other reinforcers are available, the present results are consistent with the hypothesis that drug self-administration by humans can be decreased by increasing the value of alternative positive reinforcers. In addition, these results suggest that the extent to which drug self-administration is sensitive to this manipulation varies across drugs.