The effect of female social-cohort groups on sex-typed behavioral preferences and conceptions of role stereotypes was studied by means of multivariate analyses. Subjects' self-reports of femininity were obtained, and May's projective technique, the Sex-role Stereotype scale, and Gough's Psychological Femininity scale were administered to a total of 105 females, ranging in age from 19 to 49 yr., who were distributed across the following groups: nursing trainees, introductory psychology students, skilled workers, and active feminists. Cohort membership was a significant factor; feminists differed from the other female groups in the direction of greater masculinity on 3 of 4 test variables and showed greater differentiation in their self-report of femininity ( p < .01). Low to moderate correlations were obtained among the test scores, age, and educational level. The problems inherent in sex-role assessment appear related to an interaction between gender-identity and cohort-influenced behavioral preferences.