The study was an investigation of the influences of mathematics achievement, mathematics self-efficacy, and life-styles orientation on university students' career choices. The first group of 249 subjects rated 83 occupations representing Holland's six general occupational themes for their compatibility with meeting leisure and relationship needs both inside and outside the job. The second group of 241 subjects identified the college major and occupation they were considering most and completed the Mathematics Self-Efficacy Scale (MSE) (Betz & Hackett, 1982) and the Life-styles Inventory - Relationship scale (LSI) (Epperson, Lucas, & Zytowski, 1983). They rated the extent of consideration they give to entering 83 occupations representing Holland's six themes. American College Tests - Mathematics (ACT-Math) scores were accessed for 166 of these subjects;Results of the first study confirmed that the three gender-stereotyped Holland themes of particular interest received relationship-compatibility ratings in descending order: Social, Investigative, and Realistic;For the second study, it was hypothesized that gender differences in ACT-Math, MSE, and LSI-Relationship would moderate gender differences in career choices. The results of hierarchical regression analyses indicated that only ACT-Math and MSE mediated gender differences in science-relatedness of major. All three mediated differences in consideration of Realistic occupations, only MSE mediated differences in consideration of Investigative occupations, and ACT-Math and LSI-Relationship mediated differences in consideration of Social occupations. The three variables did not predict differentially for males and females. Post-hoc analyses suggested that gender differences in LSI-Relationship may affect entry into physical sciences.