The effect of practice and training in spatial skills on scores obtained by male and female students on the Embedded Figures Test was examined. Forms A and B were administered 6 wk. apart to three groups of subjects (ns = 28, 27, 27) enrolled in drafting, mathematics, and liberal arts courses. During the pretest-posttest period the drafting students received training while the other two groups served as controls. Analysis indicated (1) no initial sex difference in test scores; (2) liberal arts students differed significantly from drafting and mathematics students, but there was no significant difference between the last two groups; (3) all groups improved with practice; (4) women receiving training improved more than women who did not; (5) there was a trend toward women receiving spatial training scoring more poorly than males receiving training on the pretest, but there was no significant difference on the posttest. These results suggest that sex differences in embedded-figures scores found by many previous experimenters may have been associated with differences in prior experience in spatial skills and by a confounding of sex with area of academic study.