Differences in tolerance and cross-tolerance to the behavioral effects of d-amphetamine sulfate and mescaline hydrochloride were studied as a function of intraventricular or intraperitoneal routes of administration. Operant behavior, using a fixed ratio schedule of food reinforcement, was the behavioral variable. Following tolerance to d-amphetamine infused intraventricularly, an approximately equipotent dose of mescaline, similarly administered the next day, resulted in behavioral disruption indicating a lack of cross-tolerance. However, following tolerance to mescaline administered intraventricularly, an equipotent dose of d-amphetamine infused the next day did not result in behavioral disruption, indicating cross-tolerance, When d-amphetamine (1.6 or 2.5 mg/kg) and mescaline (10 mg/kg) were administered peripherally (i.p.), no cross-tolerance was found, regardless of the order of drug presentation. Finally, tolerance to amphetamine administered intraventricularly was not indicative of tolerance to amphetamine given intraperitoneally. The next day (after central tolerance) amphetamine disrupted the operant to an equal or greater extent than it had prior to central tolerance formation. This is further evidence that drugs administered directly into the CNS may not be producing their behavioral effects in the same manner as they do via peripheral administration.