The purpose of this research was to test whether delay between a mug book task and a lineup task moderates a simple mug shot exposure effect. Following the witnessing of a simulated theft, participants searched either a small, large or no mug book. Participants then viewed a lineup that contained either the perpetrator or an innocent replacement; the lineup occurred either immediately after viewing the mug book or 48 hours later. There were fewer suspect identifications and more lineup rejections in the mug book conditions than the no mug book control; this was explained in terms of the criterion for making choices carrying over from the mug book to the lineup. There was qualified support for delay moderating the mug book exposure effect. Copyright © 2009 John Wiley & Sons, Ltd.