Numerous studies of retrieval-induced forgetting have shown that the selective retrieval of some studied items can impair recall of other nonretrieved items. Varying the lag between study and selective retrieval and using lists of unrelated items as study material, recent work replicated this detrimental effect when the lag between study and selective retrieval was short but reported a beneficial effect of selective retrieval when the lag was long. Here we report the results of 4 experiments in which we examined the influence of lag (3 min vs. 24 hr) for the effects of selective retrieval in categorized lists. When the selectively retrieved and the nonretrieved items shared the same categories (Experiments 1 and 2), we found detrimental effects of selective retrieval regardless of lag between study and selective retrieval. In contrast, when the selectively retrieved and the nonretrieved items belonged to different categories (Experiments 3 and 4), the effects of selective retrieval varied with lag, showing a neutral effect after the short lag and a beneficial effect after the longer lag. A 2-factor explanation is provided that assumes critical roles in selective retrieval of (a) inhibition and blocking and (b) context retrieval. This account captures the present findings as well as the recent results on the effects of selective retrieval with lists of unrelated items.