This overview examines recent research and thinking about school size and bases discussion on two key assumptions: good schools can differ widely in size, and there is no such thing as "optimal" school size. Rather, the "right" size for a school depends on local conditions and contexts. School size means more than total student enrollment. Grade span and grade level must be taken into account, and enrollment-per-grade is a more useful measure of size. In 1997-98, school enrollment in the United States ranged from 4 students in a Nebraska K-6 school to 5,160 students in a Florida 9-12 high school. Precise definitions of "large" and "small" do not exist, and the range of "informed judgments" about upper limits is substantial. Other context factors related to the effects of school size are socioeconomic status, rural versus urban locale, and state policies. Research is summarized concerning the influence of school size on student achievement, equity of achievement across socioeconomic levels, extracurricular participation, school climate, and dropout rates. This research indicates that school size alone can have a positive or negative influence on achievement levels, but that small schools are more effective in impoverished communities and make achievement dramatically more equitable. Secondary interpretations of the literature suggest that many schools are too large to serve students well, and smaller schools are widely needed, particularly in impoverished communities. Advice is offered to administrators in the form of principles to guide decision making about school size, policy mechanisms that support small schools, recommended local action, and the challenges and benefits of house plans and schools-within-schools. (Contains 61 references and 5 suggested readings.) (SV)