Four experiments were conducted to explore the effects of prior exposure to word stimuli on a) identification of the words under perceptually impoverished conditions, and b) recognition of the words as having been previously presented. In Experiment 1, distributing a two-second study duration between two one-second or four-second presentations as opposed to concentrating it into a single two-second presentation was found to enhance perceptual identification but have no reliable effect on recognition. Experiment 2 showed that changing modalities between study and test presentations (i.e. from visual to auditory or auditory to visual) reduces but does not eliminate the effect study presentation has on perceptual identification. Experiments 3 and 4 demonstrated that identification both of word-fragment cues and of tachistoscopic stimuli declines sharply over very brief study-to-test intervals but then stabilizes for intervals of at least 24 hours.