Participants tend to initiate immediate free recall (IFR) of short lists of words with the very first word on the list. Three experiments examined whether rehearsal is necessary for this recent finding. In Experiment 1, participants were presented with lists of between 2 and 12 words for IFR at a fast, medium, or slow rate, with and without articulatory suppression (AS). The tendency to initiate output with the first item for short lists (a) did not change greatly when presentation rate was increased from a medium to a fast rate under normal conditions, (b) was reduced but not eliminated by AS, and (c) was maintained at slower rates when rehearsal was allowed but decreased at slower rates when rehearsal was prevented. In Experiment 2, the overt rehearsal methodology was used, and the tendency to initiate output with the first item for short lists was present even in the absence of overt rehearsal. Experiment 3 re-examined IFR under normal encoding conditions and replicated the main findings from the normal encoding conditions of Experiment 1 while using the presentation rates and list lengths of Experiment 2. We argue that rehearsal is not strictly necessary for the tendency to initiate recall with the first item under normal conditions, but rehearsal nevertheless contributes to this effect at slower rates. (Contains 3 tables and 14 figures.)