The conditioned place preference paradigm was used to investigate the reinforcing properties of cocaine. Rats were injected (i.p.) with cocaine hydrochloride (0.625-20 mg/kg) and then immediately confined for 30 min to one side of a shuttle box in which each of the two compartments had distinctive features. On alternate (control) days they received saline injections and were confined for 30 min to the opposite side. Cocaine produced a significant, dose-related preference for the distinctive environment that previously had been paired with the drug. Pretreatment with pimozide (0.5 or 1.0 mg/kg) or haloperidol (1.0 mg/kg), both of which blocked the locomotor stimulant effects of cocaine, failed to influence place preference produced by cocaine (5.0 mg/kg). In addition, neither 6-hydroxydopamine lesions of dopamine terminals in the nucleus accumbens nor 6-hydroxydopamine-induced destruction of central and/or peripheral noradrenergic systems affected cocaine-induced place preference conditioning. In other experiments it was found that injections of the local anaesthetic procaine, at doses that did not affect locomotor activity (25 and 50 mg/kg), also resulted in significant place preference conditioning. It is concluded that cocaine can produce place preference conditioning through a mechanism that is independent of its effects on catecholamine-containing neurons and that may be related to its local anaesthetic properties. It is noted, however, that if cocaine's local anaesthetic properties could be blocked selectively, the drug might still produce place preference conditioning through its enhancement of central dopaminergic function.