In this study it is found that individual school buildings (the complex human social systems found in schools, not the buildings themselves) do affect student performance and achievement. A grade equivalent achievement test (total reading and total math score only) was used to measure achievement of 2,500 Detroit elementary school children in the 1972-1973 school year. Through the use of a multiple regression model, it was found that: (1) previous achievement is significantly predictive of current achievement; (2) knowledge of school attended accounts for about 22% of the variance in reading and 21% of the variance in math test scores; (3) schools are not simple proxy variables for the socioeconomic background of students they serve; and (4) the amount of variance accounted for in reading and math achievement is significantly increased when knowledge of individual school attended is included with prior measures of achievement. Four tables substantiate the findings. It is suggested that educational policy makers take into account research which demonstrates the importance of individual school characteristics, such as positive staff attitudes, strong administrative leadership and instructional strategies which have been proven effective in urban schools with populations of low socioeconomic status. (WI)