The present study investigated the effects of presentational timing, operationalized as different levels of temporal overlap, on cross-modal priming of written words. We used a paradigm where the playback of spoken word primes was shifted relative to the presentation of written targets (asynchronous, partially overlapping, and synchronous presentation). Our participants (n = 48) carried out a speeded lexical decision task on the written targets. Presenting the spoken primes, albeit the words' onset, before the written targets reduced lexical decision times to both words and pseudowords. Asynchronous presentation of the spoken primes resulted in the largest difference between word and pseudoword response times. We discuss our results in relation to the mental structure of human word knowledge and in the context of word form acquisition.