The focus of this paper is to further understanding of the distribution and the effects of an expanded conception of opportunity to learn on student mathematics achievement. In addition to descriptive statistics, a set of two-level hierarchical linear models was employed to analyze a subset of the restricted-use National Education Longitudinal Study of 1988 database. The results revealed that on different scales, various kinds of opportunities to learn mathematics are associated with student mathematics achievement, and opportunities are unequally distributed among different categories of schools. Four implications for educational policymaking are provided. They are: (1) the need to recruit, retrain, and retain teachers with adequate mathematical knowledge; (2) to encourage high content and level of instruction (including a high level of instruction, broad coverage, and an optimum amount of homework, which should be meaningful rather than merely busywork); (3) to provide more advanced mathematics courses, including replacing general mathematics with more advanced courses for the majority of the student body without decreasing the content; and (4) to increase learning opportunities in disadvantaged areas. Contains 85 references and 4 data tables. (Author/PVD)