The paper describes problems endemic to the evaluation of programs for gifted and talented students. The goals of gifted and talented programs are often substantially different from those covered on outcome measures available to schools. When decisions are being made about the type of programs to be provided, parallel consideration should be given to how the program is to be evaluated. Two dimensions on which gifted and talented programs generally vary are acceleration vs. enrichment and in-class provision of instruction vs. pull-out programs. Examples are then given of programs which follow each of these four models: in-class enrichment, in-class acceleration, pull-out enrichment, and pull-out acceleration. Characteristics of good evaluation consider 30 standards organized into the following four categories of quality: utility, feasibility, propriety, and accuracy. A sampling of these standards as they apply to gifted and talented programs is discussed. These standards include: audience identification, information scope and selection, valuational interpretation, practical procedures, rights of human subjects, object identification, valid and reliable measurement. It is concluded that the evaluation of gifted programs requires more care in the identification of audience, the lack of common goals for all students, and lack of appropriate instruments. (DB)