This study assessed patterns of differences in quantitative performance across groups of intended undergraduate majors consistent with those previously found for students who had completed their undergraduate study. Data were drawn from the College Board Admissions Testing Program's national sample of 10,000 college-bound high school seniors in 1982-83. The data base includes student scores on the Scholastic Aptitude Test (SAT) as well as student reported information from the Student Descriptive Questionnaire, which elicited information on students' family background, high school experiences, personal characteristics, and educational aspirations. The model for the study was estimated for 314 women who indicated their intention to major in one of the mathematical or scientific fields within the scope of the study. Women intending to major in engineering and physical science exhibited the same atypical performance. However, the large differences in mean SAT Mathematics performance found between women in mathematics, statistics, physics, and computer science versus those in engineering and the physical sciences disappeared when other measures were controlled for; and the effects of these measures were the same for both groups. Self-rating relative to others with respect to mathematics and science ability was the primary influential variable in the study. (TJH)