Two experiments were undertaken in which the effects of semichronic administration of the precursor steroid, pregnenolone, were examined in a food search task. In both experiments male rats were required to find a food reward in a designated hole in an arena with 16 equally spaced holes. Hormone administration began 8 days before the onset of training. Training was given on an every-other-day schedule for five sessions. Animals were deprived of food for 18 h before training or testing. Retention testing occurred 10 days after acquisition and this was followed by 2 days of training using a different hole for the reward. The two experiments differed only in the method of hormone administration. In one experiment the rats received an implanted (sc) slow release pellet containing pregnenolone before training. In the second experiment the animals received ip injections of pregnenolone sulfate before and during initial training and then had the slow release pellet implanted between acquisition and retention. Significant enhancement of retention was found during the middle trials of the retention test when the treated and control groups from the two experiments were combined. No differences were found during acquisition training in either experiment. On the first day of training the animals to find the reward in a new location, the group injected with pregnenolone sulfate and later implanted with pregnenolone slow-release pellets exhibited performance superior to that of their matched control group.