Experimental data on memory ability of autistic children is scarce and contradictory. It is suggested that this may in part be the result of failure to control for acquisition of the correct response before memory is tested. In the first experiment reported, acquisition was controlled in a delayed-response visual discrimination task. The performance of autistic children was compared with MA matched normal and retarded controls. Recall deteriorated with increasing delay interval in each group. There were no significant group differences. The second experiment was designed to measure the effects of interference on memory using a serial memory task. No differences were found among groups in rate of learning, although the autistic children improved significantly on the second and longer series. There was no relationship between performance on the STM and the serial memory tasks over the same time intervals for any group. The results are interpreted as supporting the view that when MA and response acquisition level are equated with normal and retarded children in a memory task, autistic children do not show any specific memory deficits.