Among the factors affecting the quality of education in schools are characteristics of principals, characteristics of teachers and programs, and characteristics of families. In this study of 80 schools in Virginia, composite fourth-grade reading and math scores, calculated by taking the mean of total reading and math scores for three years (83-85), were regressed on three sets of variables representing these factors. Each set of variables was entered as a block, and all orders of entry were systematically explored. When characteristics of families were entered first, they accounted for 56 percent of the variance in reading scores and 35 percent of the variance in math scores. The change in R² was not significant when each of the remaining blocks were entered. Order of entry had no effect; characteristics of families were the only variables that produced a statistically significant R square, regardless of order. Both variables comprising characteristics of families produced significant partial regression coefficients. The coefficients for percentage of students in the school receiving free or reduced-price lunch were -.632 (<u>p</u>< .00) for reading and -.539 (<u>p</u> < .00) for math. The coefficients for percentage membership in the school‘s organization for parents were .077 (<u>p</u>< .05) for reading and .101 (<u>p</u> < .05) for math. Characteristics of principals and characteristics of teachers and programs, as measured in this study, did not have an effect on student achievement in reading or math above their combined effects with characteristics of families. Obviously, principals, teachers, and programs do have an effect on student achievement; however, the specific variables used in this study to measure these characteristics were unable to capture these effects. / Ed. D.