The relative role of the timing and repetition of misinformation on the accuracy of children's recall was examined in two experiments. Kindergarten children participated in a magic show and about 40 days later had a memory test. Between the magic show and the memory test, the children were suggestively interviewed either one time in a relatively ‘early’ interview (temporally closer to the magic show than the memory test) or a relatively ‘late’ interview (closer to the memory test than the magic show), or in both suggestive interviews. The timing of the suggestive interviewing was manipulated so that the interview was temporally distant from the event or memory test or temporally close to the event or memory test. Repeated interviewing heightened misinformation effects only when the children received the two interview sessions temporally close to the event and memory test. Copyright © 2004 John Wiley & Sons, Ltd.