This paper discusses the many failures of comprehensive public schools in urban areas and looks at possible alternatives. It uses data from the National Educational Longitudinal Survey to determine the academic benefits, if any, of attending a magnet school, Catholic school, or secular private school.