The major research question addressed in this dissertation is: To what extent do past attainment models which have been primarily applied to white male samples predict the educational achievement of blacks and women? Data for this inquiry are based on a subsample of black and white males and females who participated in a National Longitudinal Survey of the high school senior class of 1972 conducted by the Educational Testing Service and the Research Triangle Institute. The analytical model used in this study is a four variable model consisting of two traditional predictors of educational attainment; family socioeconomic status and mental ability; an intervening variable (high school senior class rank); and the major outcome variable, educational achievement as measured by college attendance. Findings from the separate parallel path analyses performed on each race by sex subgroup supported the proposition that the educational achievement process itself varied more by race than by sex. As for the direct effects of race and sex on educational achievement when entered explicitly into the basic model, the results from path analysis were contrary to the effects of race hypothesized in that being white produced a significant direct depressant effect on educational achievement rather than being black. However, as was hypothesized, being female did produce a modest direct depressant effect on the major outcome variable. Thus a major conclusion drawn from this study is that the presence or absence of race and sex differences in the educational achievement process is primarily contingent upon the influence of socioeconomic status and ability. (Author/AM)