Drawing on Eccles' expectancy-value model of achievement-related choices, we examined the personal aptitudes and motivational beliefs at 12th grade that move individuals toward or away from science, technology, engineering, and mathematics (STEM) occupations at age 29. In the first set of analyses, occupational and lifestyle values, math ability self-concepts, family demographics, and high school course-taking more strongly predicted both individual and gender differences in the likelihood of entering STEM careers than math scores on the Differential Aptitude Test. In the second set of analyses, individual and gender differences in career decisions within STEM disciplines (health, biological, and medical sciences (HBMS) versus mathematics, physical, engineering, and computer sciences (MPECS)) were best predicted by occupational values (i.e. preferences for work that were people oriented and altruistic predicted entrance into HBMS instead of MPECS careers). Females were less likely to hold the beliefs that predicted selection of STEM in general, but those who did choose STEM were more likely to select HBMS than MPECS. One Sentence Summary: Gender differences in selecting STEM related and health, biological, and medical occupations result primarily from gender differences in occupational and lifestyle values.