In separate experiments, rats received either five intraperitoneal or five subcutaneous injections of cocaine (once daily or spaced every fourth day) prior to receiving repeated saccharin-cocaine pairings (during taste aversion conditioning). Both spaced and massed subcutaneous cocaine preexposure attenuated the subsequent acquisition of taste aversions induced by cocaine. Specifically, aversions in the preexposed subjects were acquired at a slower rate and/or to a lesser degree than those acquired by subjects preexposed to the cocaine vehicle and injected with cocaine during conditioning. Spaced and massed intraperitoneal cocaine preexposure had only a weak or no effect, respectively, on the subsequent acquisition of cocaine-induced taste aversions. Specifically, subjects receiving spaced intraperitoneal injections of cocaine during preexposure differed from nonpreexposed subjects on only a single conditioning trial, and subjects receiving massed intraperitoneal injections of cocaine during preexposure displayed aversions comparable to those of nonpreexposed subjects. Although the effects of subcutaneous cocaine preexposure were similar to those reported with other drugs within the aversion design, it is clear that the preexposure effect with cocaine is dependent upon the specific parameters of preexposure. Several possibilities for these differential effects of cocaine preexposure were discussed, including the influence of changing the route of administration from preexposure to conditioning (i.e., from IP to SC) and the differential masking of the aversive effects of cocaine during conditioning by differential sensitization to cocaine's reinforcing properties following SC and IP preexposure. Although the present series of experiments did not directly address the mechanism(s) underlying the attenuating effects of cocaine peexposure on aversion learning, several possibilities were noted, including adaptation or tolerance to the aversive effects of cocaine and sensitization to its rewarding effects. (C) 1998 Elsevier Science Inc.