The tip-of-the-tongue (TOT) phenomenon was explored in light of its theoretical implications for models of lexical access. In particular, the transmission deficit theory of TOTs was tested by observing the effects of word frequency and recent word exposure on TOT incidence. TOTs were induced in 156 participants by asking them to retrieve both high and low-frequency words from definitions. Participants had earlier completed an ostensibly unrelated task in which half the target words (recency manipulation) were presented in either phonological or semantic rating tasks. A transmission deficit model would expect this task manipulation to show no exposure effect on TOTs following the phonological task, and an exposure by frequency interaction following the semantic task. The results tended to concur with these expectations, but there were some notable exceptions. Closer examination of the subject data revealed effects tending to vary with the age of participants. Examination of the item data suggested that lack of an exposure effect on TOT incidence might not be absolute, but reflective of no net change in TOTs. Results are discussed in terms of the hypothetical mechanisms underlying lexical retrieval and implications for the locus of TOTs.