To gather data on the implications of the proposition that intellectual ability should be conceived as multidimensional, 19 different measures, all of which have been employed as selection criteria for programs for the gifted, were used as a basis for selecting 5 students from 2 fifth-grade classes (N=49) for inclusion in a hypothetical program for the gifted. The principal hypothesis, that different students would be selected as a function of the selection technique utilized, was supported. No student was selected by all techniques, seven students were selected by only one technique, and ninety-two percent of the children (N=44) were selected on the basis of one or more technique. It is concluded that communities should be informed of the implications of various selection techniques so that each may be judged on its merits when establishing special programs for talented students. (Author)