Four groups of five food-deprived hooded Long-Evans rats were injected subcutaneously with saline (vehicle) or 2, 4 or 8 mg phencyclidine (PCP) hydrochloride/kg fifteen minutes before being placed for the first time into operant chambers modified to detect exploratory behaviors. Rearing was found to be more sensitive to disruption by phencyclidine than was unconditioned level touching (a measure of floor-level exploratory activities). In an autoshaping session immediately following, the group of animals given the low dose of PCP made as many lever-touch responses as the group given saline, but consumed fewer of the food pellets delivered. In addition, none of the animals in the low-dose group showed within-session shortening of the latency to respond which was observed in four of five control animals. The two other groups given higher doses of PCP demonstrated dose-related decrements in responding as well as a reduction in food pellet consumption during the first session of autoshaping. Over the next two daily autoshaping sessions, performance improved in those groups initially suppressed. Performance converged in all group by the third autoshaping session.