For the past twenty years assessment instruments have indicated a relationship between personality variables and occupational choice. Since males and females tend to have personality traits and needs that differ from each other, sex differences are also important factors in occupational choice. Moreover, the study of sex differences involves the effects of testosterone on behavior. Testosterone has a known effect on the brains of animals thereby influencing aggressive and sexual behavior. It is possible that testosterone is related to human behavior and personality as well. The present study investigated the relationship in women between testosterone concentration, selected personality needs and traditional and non-traditional occupational choice. Personality needs, as measured by the Edwards Personal Preference Schedule (EPFS), included four selected masculine needs - achievement, autonomy, dominance, aggression, and two selected feminine needs - deference and nur- turance. Elementary school teaching and nursing represented traditional occupational choice, while professional athletics and law represented non-traditional occupational choice. Three hypotheses were developed for the study. First, significant positive relationships were hypothesized between testosterone concentration and each of four selected EPPS masculine needs, and significant negative relationships were hypothesized between testosterone concentration and each of two selected EPPS feminine needs. Also, significant mean differences were hypothesized between the four occupational groups in terms of testosterone concentration and the six selected EPPS needs. A last hypothesis dealt with whether testosterone concentration and scores on the six selected EPPS scales could be used to classify the subjects into occupational groups. The sixty-four women volunteering for this study were comprised of four equal groups: suburban school district elementary teachers; urban clinic nurses; professional team sports athletes; urban lawfirm attorneys. All subjects were under the age of forty-three; neither menopausal nor pregnant; had at least one functioning ovary; and smoked no more than two marijuana cigarettes daily. The measures included the EPPS administered under controlled conditions and a measurement of testosterone concentration in saliva collected by each subject on Day 6 of her menstrual cycle. Salivary testosterone concentration is significantly correlated to free serum testosterone concentration (r = .83, p<.001). The newly developed, non- invasive salivary technique allowed for routine, painless collection of biological fluid. [...]