This paper presents a fixed effects model on the impact of individual high schools on adult earnings. This model not only captures the effects on observed and unobserved characteristics of high school on earnings, but also can control for selection bias resulting from correlation between school quality and student ability. The principal data used to estimate this model comes from the National Longitudinal Study of the High School Class of 1972. The sample for the analysis was restricted to out-of-school white males and females who lived in Chicago, Detroit, Los Angeles, New York, Philadelphia, or San Francisco. The findings suggest that such a model has a large impact on estimates of school quality and, as a result, on the policy conclusions that are drawn on the basis of such evidence. This report first describes a general model of the effects of school quality and outlines the conditions required to obtain consistent parameter estimates. Second, it points out the problems of previous work in conforming to these conditions. Third, it describes the approach used to estimate school quality. Finally, it presents empirical results. (PN)