The authors consider the problem of classifying an unknown observation into 1 of several populations by using tree-structured allocation rules. Although many parametric classification procedures are robust to certain assumption violations, there is need for classification procedures that can be used regardless of the group-conditional distributions that underlie the model. The authors discuss the tree-structured allocation rule. The tree rule is compared with linear discriminant and logistic regression procedures via Monte Carlo and real data analysis. (Contains 6 tables, 1 figure, and 1 note.)