Much research on sex roles and their correlates characteristically has been performed in laboratory settings with a college student sample. Job analysis interviews were conducted to extend androgyny research to the work setting. Supervisory performance ratings of job activities and Bem Sex-Role Inventory data were obtained for 135 female nurses. A comparison of masculinity and femininity ratings of the job activities resulted in five tasks being classified as "stereotypic feminine" and four as "stereotypic masculine." Findings indicated that the masculine sex role stereotype was relevant to the prediction of job performance for the stereotypic masculine tasks. Generally, prediction of performance for the feminine tasks was not obtained. Evidence was not present for the notion that the concept of androgyny can aid in the prediction of performance in both stereotypically masculine and feminine job activities. (Author)