Using eighth-grade transcript and questionnaire data from the National Education Longitudinal Study of 1988, this study used logistic regression procedures to investigate the extent to which the probability of mathematics proficiency depends on the values of various background factors. The results of the analyses, which were based on a subsample of approximately 450 students, indicate that when race and gender were the sole explanatory variables, the probability of mathematics proficiency at the highest level was greatest for Asian/Pacific Islanders and Whites. Using this model, the odds of mathematics proficiency for females were no different from that of males. When other factors were included in the model, the probability of mathematics proficiency increased with advanced math course enrollment, grades, socioeconomic status, reading proficiency level, and homework. In addition, when other factors were controlled, males were more likely to achieve at the highest level of proficiency than females. A cumulative logit model was used to assess the effect of the explanatory variables on the cumulative probabilities for math proficiency. The results of this analysis reveal that attendance at private nonreligious schools, advanced math enrollment, and increased homework increase the odds of higher mathematics proficiency. Chi-squared tests of the association between mathematics proficiency and the covariates as well as correlation analyses among the covariates suggest that the absence of some variables from the prediction equations may be a results of multicolinearity among the explanatory variables. (Author/SLD)