In response to recent changes in the Home Office recommendations concerning police interviewing of mentally handicapped persons, an experiment was carried out to investigate the usefulness of different interviewing techniques with mildly mentally handicapped children. Previous research carried out with children of normal intelligence (Dent&Stephenson, 1979) indicated that their recall of an incident was most accurate when unprompted. Theoretical research in the field of mental handicap suggested that such children's recall would be poorest when either unprompted or when heavily prompted, and that some form of intermediate cueing of recall may prove optimal. The experiment described here investigated the accuracy of recall of a live incident by a group of children with IQs ranging from 50 to 70 points in response to one of the following methods of elicitation: free recall, general questions and specific questions. As predicted, the general questions produced recall that was optimal in terms of completeness and accuracy.