Despite accumulating evidence of gender differences in mental rotation performance, much remains unknown about the variables that lead to an advantage among boys compared to girls. Here we examined the role of trial history on children’s performance. To this end, we manipulated the difficulty of trials and implemented drift diffusion modeling (DDM) to assess how prior exposure to easy versus hard trials affects the parameters of drift rate and decision threshold. In Experiment 1, children were presented with either an easy-to-hard or a hard-to-easy block order. On easy trials, there were no gender differences in accuracy or drift rates, regardless of order. On hard trials, girls matched boys in accuracy and drift rates, but only when easy trials were presented first, suggesting that girls’ performance on hard trials benefited from prior exposure to easy trials. In Experiment 2, we ruled out a general practice effect, confirming that improvement in girls’ performance is specific to exposure on easy trials prior to hard trials, not just more trials. Additionally, boys, in general, had larger decision thresholds than girls. Taken together, these findings point to gender differences in mental rotation performance that are dependent on trial history and that may reflect differences in affective and/or motivational factors between boys and girls.