Social-learning theories of substance use propose that members of peer groups influence the drug use of other members by selectively modeling, reinforcing, and punishing either abstinence-related or drug-related behaviors. The objective of the present study was to examine the social influences on cocaine self-administration in isolated and socially housed rats, under conditions where the socially housed rats were tested simultaneously with their partner in the same chamber. To this end, male rats were obtained at weaning and housed in isolated or pair-housed conditions for 6 weeks. Rats were then implanted with intravenous catheters and cocaine self-administration was examined in custom-built operant conditioning chambers that allowed two rats to be tested simultaneously. For some socially housed subjects, both rats had simultaneous access to cocaine; for others, only one rat of the pair had access to cocaine. An econometric analysis was applied to the data, and the reinforcing strength of cocaine was measured by examining consumption (i.e. quantity demanded) and elasticity of demand as a function of price, which was manipulated by varying the dose and ratio requirements on a fixed ratio schedule of reinforcement. Cocaine consumption decreased as a function of price in all groups. Elasticity of demand did not vary across groups, but consumption was significantly lower in socially housed rats paired with a rat without access to cocaine. These data suggest that the presence of an abstaining peer decreases the reinforcing strength of cocaine, thus supporting the development of social interventions in drug abuse prevention and treatment programs.