Research across seven very different states has shown that as schools get larger, the average achievement among economically disadvantaged students declines. A traditionally strong argument against smaller schools, however, is that they are too expensive. Large consolidated schools with narrowly specialized grade spans are typically offered as necessary to save money and to meet the needs of different age groups. This paper examines whether the size-by-socioeconomic status interaction effect proves robust across alternative regression model specifications, as it did across different states. It also explores whether the equity gains associated with smaller schools are incompatible with the need to save money. Data were analyzed for 1,001 Texas high schools, including 116 single-unit schools--K-12 schools that are the only school in their district. Analysis with four different regression model specifications consistently found that increased school size was related to increases in achievement-test-score costs associated with economically disadvantaged students. School size was negatively related to expenditure per pupil, but this relationship became increasingly tenuous as size increased, and eventually savings became negligible. In addition, organizational factors, especially as manifest in the distinctive components of the single-unit school, were related to cost reduction. For those interested in balancing expenditure per pupil with achievement-based equity, the best grade-span configuration seems to be a small single-unit school. (Contains 76 references.) (SV)