Latent inhibition (LI) is defined as poorer evidence of learning with a stimulus that previously was presented without consequence, as compared with a novel or previously attended stimulus. The present article reviews the evidence, mostly from three-stage conditioned taste aversion studies (preexposure, conditioning, and test), that LI can be either attenuated or enhanced depending on the length of the retention interval between conditioning and test and where that interval was spent. Time-induced reduction in LI is observed when the interval context is the same as that of the preexposure, conditioning, and test stages. Super-LI is obtained when a long retention interval is spent in a context that is different from that of the other stages. The differential modulations of LI appear to be the result of the strengthening of primacy effects (i.e., first training disproportionately stronger than subsequent training) by long-interval different contexts, thereby producing super-LI, and the reversal of this effect by long-interval same contexts, thereby producing attenuated LI. The bidirectional effects of time/context modulations on LI, unaccounted for by current learning theories, are explained, in part, by a time-induced context differentiation process. Implications for theories of LI, learning, and memory are discussed.