From a sample of 6,000 pupils in a Tennessee school district, this study relates the variables of gender, social class, and number of parents in student households to achievement on a norm-referenced test. Charts presenting data findings comprise three-fourths of the document. Investigation was motivated by studies that connect student learning difficulties to poverty and by research claims that students from all social classes can master minimum competencies. Researchers disaggregated California Achievement Test (CAT) data over a 3-year period on the basis of social class from pupils in grades 4, 6, and 8. Third- year data were additionally disaggregated by gender and by number of parents. Results show that upper social class pupils performed better than lower social class pupils, though that difference narrowed during the three-year period. On the CAT, better scores were produced by females, two-parent children, and upper social class children. Two-parent lower social class pupils performed better than did their counterparts from one-parent households. Conclusions are as follows: (1) identifiable social class subgroups of students are not performing well; (2) this study should be replicated; (3) because poverty affects educational outcomes, schools require assistance; and (4) curriculum revisions are needed. (CJH)