Current evidence about the persistence of syntactic priming effects (Bock, 1986) is equivocal: Using spoken picture description, Bock and Griffin (2000) found that it persisted over as many as 10 trials; using written sentence completion, Branigan, Pickering, and Cleland (1999) found that it dissipated if even a single sentence intervened between prime and target. This paper asks what causes it to be long lasting. On one account, the rapid decay evidenced by Branigan et al. occurs because the task emphasizes conceptual planning; on another account, it is due to the written nature of their task. If conceptual planning is the cause, this might relate to planning the prime sentence or planning an intervening sentence. Hence we conducted an experiment with spoken sentence completion, contrasting no delay, an intervening sentence, and a pure temporal delay. The results indicated that strong and similar priming occurred in all three cases, therefore lending support to the claim that spoken priming is long lasting.