Accurate classification in discriminant analysis and the value of prediction are discussed, with emphasis on the uses and key aspects of prediction. A brief history of discriminant analysis is provided. C. J. Huberty's discussion of four aspects of discriminant analysis (separation, discrimination, estimation, and classification) is cited. Predictive accuracy results when an investigator understands certain main rules and validation methods. The generalizability of discriminant analysis results, and external versus internal analysis are considered. Four traditional types of external validation methods are discussed: (1) the empirical method; (2) the holdout method; (3) Monte Carlo method; and (4) the random assignment method. Two non-traditional methods, the jackknife and the U-method, are also reviewed. Focus is on the U-method, which is illustrated using a hypothetical data set of 64 cases for which the actual classifications (four groups) are known. Four classification tables are provided, which demonstrate concepts such as hit rates, leave-one-out, and predictor ordering. A summary is presented to improve the interpretation of discriminant analysis results and multivariate procedures in general. A 15-item list of references is included, and selected file commands are outlined. (SLD)