Objective: To explore the possibility that gender has a moderating effect on memory after pediatric traumatic brain injury (TBI). Design: Controlled group study. Gender effects between and within groups were evaluated by means of effect size comparisons and hierarchical regression analysis. Setting: Regional rehabilitation center. Participants: Seventy children with TBI, selected from a 4-year series of consecutive referrals, and 70 demographically matched controls. Main outcome measures: Screening version of the Wide Range Assessment of Memory and Learning (WRAML-S) and the Wechsler Intelligence Scale for Children-Third Edition (WISC-III). Results: Boys with TBI performed worse than girls with TBI, and worse than their counterparts in the control group, on the WRAML-S. There was no gender effect in the control group. Gender explained an additional 9% of the variance in WRAML-S performance over and above injury severity and age variables. However, gender differences were largely attenuated when speed of information processing, as assessed by the WISC-III, was used as a covariate. Conclusion: The effect of TBI on children's memory appears to be moderated by gender and may be mediated by speed of information processing.