This study investigated the effects of contextual interference on the immediate transfer and 2-day retention of a bean bag tossing task by mildly mentally handicapped children. A total of 24 boys and 24 girls with a mean chronological age of 10.2 years were randomly assigned to either a blocked, serial, or random practice condition. Following 48 practice trials with bean bags of various weights, subjects were transferred to two novel weighted bean bags. Both transfer and retention analyses showed that subjects in the random practice condition exhibited less error than subjects in either the blocked or serial practice conditions. However, these differences were not significant. Boys performed with significantly less error than girls on both transfer and retention, while regardless of gender, the heavier weighted bean bag resulted in significantly less error on transfer only. Results provide marginal support for the contention that greater contextual interference (random practice) leads to better transfer and retention than other types of practice conditions.