In the current educational climate, a great deal of emphasis is placed on measuring both teacher and school effectiveness. We employed a residual based approach to examine school effectiveness using school mean achievement data from one state. Rather than evaluating schools based on their mean achievement, we chose to predict achievement based on a variety of school demographic and background variables and identified schools that had achievement profiles that exceeded or fell short of their expected achievement levels. Next, we conducted a series of correlational and latent variable analyses to explore the structure of the residuals across time and across subject area. The results demonstrate that the school residuals clearly contain a great deal of signal, as well as noise.