The purpose of the present study was to investigate the false memory rates for word and picture Deese-Roediger-McDermott (DRM) lists in children, adolescents and adults with autism spectrum disorder (ASD). Deese (1959), Roediger and McDermott (1995) developed the DRM paradigm as a method for testing the effects of semantic intrusion on the creation of false recollections. This method is used in the current thesis to test false memory in participants with Autism Spectrum Disorder (ASD), a disorder characterised by social, language and behavioural deficits. False memory studies of adults with this disorder have had conflicting results and found ASD participants have lower or similar false memory rates to controls. Experiment 1 compared false memory rates in 11 children, 11 adolescents and 5 adults with ASD to 15 children, 11 adolescents and 7 adult controls. ASD participants had higher false critical lure rates and lower studied item recognition rates than controls regardless of age. Adolescents had higher false and studied memory than children while adults had higher studied but lower false recognition rates than the younger groups. Due to the use of short DRM lists the adult participants recognised the fewest critical lures. In Experiment 2 there were 6 ASD children and 9 ASD adolescents compared with 6 control children and 9 control adolescents who were all tested both individually and in collaborative trios. Collaboration was beneficial to ASD adolescents and control children and adolescents by reducing false recognition and increasing studied item recognition. Collaboration was more beneficial for ASD adolescents and control children in the correct rejection of critical lures and for control adolescents in the recognition of studied items possibly due to decision making techniques. Critical lure recognition did not vary between collaborative trios and individuals and studied item recognition was decreased by collaboration in the ASD children. The observed decision making techniques employed revealed a clear pattern in social development and suggest ASD adolescents, but not ASD children, would benefit from group work. Overall the findings of this study, when taken into context with previous DRM studies, suggest that ASD participants have a delay in their social development and in the development of their association networks.