Presents and discusses the results of a study undertaken at Washington State University to determine whether different subgroups of the library collection had different loss and misshelving rates and, if so, whether the worst problems could be predicted from circulation data. Two tables present the data from the study. (Author/JL)