Two experiments were carried out to compare the performance of male and female students at different educational levels on tasks that required mental rotation. Exp. 1 also compared their performance on an overt, male-typed version and a disguised, female-typed version of the same task. Amongst introductory undergraduate students, men performed significantly better than women, but this difference was as pronounced on the disguised, female-typed version as on the overt, male-typed task. However, there was no sign of any gender difference on the overt task in advanced undergraduate and postgraduate students. The latter finding was not replicated in Exp. 2, in which significant effects of gender regardless of the students' educational level were noted. Nevertheless, the effect size was significantly smaller than that obtained for comparable students tested on the same task during the 1970s. Taken together, these results confirm that gender differences in at least some aspects of mental rotation may be abolished by educational experience and that gender differences in mental rotation have become smaller over the last 20 years. Such findings favor sociocultural explanations of gender differences in mental rotation rather than biological explanations.