In experiment 1, a CS-UCS interval study of place conditioning and analgesic tolerance with morphine was conducted. Morphine (10 mg/kg i.p.) was administered to separate groups of rats either 2 h prior to, 1 h prior to, immediately prior to, immediately after or 2 h after 30-min confinement in one end compartment of a place conditioning apparatus. A total of three choice tests was given, one after every six morphine injections. A preference for the end compartment contingent upon morphine injection was shown in groups that received morphine prior to end compartment placement. Groups that received morphine after end compartment placement were not different in their preference behavior from groups that received only saline during place conditioning training. A hot-plate test for tolerance to the analgesic effect of morphine was given at the end of all choice testing. All groups that had received morphine during place conditioning training were equally tolerant. These results indicate a dissociation between the analgesic effect of morphine and the effect that produces place preference, since the former was not affected by temporal parameters that did affect the latter. In the second experiment, the effect of extinction on a morphine-induced place preference was studied using extinction procedures that, in contrast to previous studies, equated exposure to both end compartments. Whereas the morphine-induced place preference was undiminished by a 10-day retention period in which animals received saline injections in the home cage, extinction trials during the same period eliminated the place preference. These results provide evidence that morphine-induced place preferences involve associative processes.