Delay-induced forgetting refers to the finding that memory for studied material typically decreases as the delay between study and test is increased. The results of 3 experiments are reported designed to examine whether this form of forgetting is primarily caused by interference effects or contextual drift effects when people engage in neutral distractor tasks during the delay. Response latency analysis was used to contrast predictions of the interference and the contextual drift view of the forgetting. The results demonstrated that prolonged delay between study and test of a list of items reduced both recall rates and mean response latencies. Because mean latency provides a reliable index of the size of people's mental search set at test, the findings suggest that prolonged delay impeded people's ability to include studied items into the search set. The results also showed that (a) mental context reinstatement before test can eliminate this effect, and (b) younger and older adults differ in their susceptibility to interference effects but show comparable delay-induced forgetting. The findings indicate that, with neutral distractor tasks, delay-induced forgetting is primarily mediated by contextual drift. Such drift reduces people's mental search set at test and, thus, decreases both recall rates and response latencies.