Formulas for estimating cross-validated hit-rates, the number of correct classifications into an a priori grouping structure, were examined. The following mathematical formulas were compared: McLachlan's formula estimator, two Snappin and Knoke smoothed formula estimators, and the analytic leave-one-out estimator. The R method was included as a benchmark, but not considered as a viable candidate because of its optimistic bias. Twenty-one data sets were taken from real classification studies from journal articles, presented papers, and textbooks. The classification methods were accomplished as specified with each data set assuming equal population covariance matrices, prior probabilities of group membership, and costs of misclassification. Results showed that in most of the data sets, the estimation methods gave very similar results. Most methods overestimated the R method hit rate with some data sets, which was not expected. The McLachlan method, which does not perform as a multiple regression formula does, gave estimates larger than that of the R-method. The leave-one-out (LOO) method appeared to be the most desirable. References and tables are provided. (GDC)