Current procedures for student evaluation of graduate programs are examined, based on three studies, and a valid measure for collecting graduate student evaluations is proposed. Study one surveyed 23 southern universities to determine the number and type of evaluations conducted. Only seven (30 percent) of the institutions conduct graduate student exit surveys or alumni surveys. Survey results are used for different purposes in each institution. including program review, self-study/accreditation, and information for the Graduate Assembly. The second study surveyed 69 graduate departments at West Virginia University. Some type of formal survey and/or interview to evaluate graduate programs is conducted by 41 departments of the university, 39 (48 percent) indicated that they do not conduct any formal type of graduate evaluation. Survey results are used for degree program and curriculum review, for department review, and self-study. A few departments use the results for faculty review. In the third study, the Graduate Student Program Evaluation test was constructed and tested. Factor analysis reduced the test to 32 items with high internal reliability. The instrument was tested with 99 graduating master's and doctoral students. The results suggest that graduate students' evaluations of their programs are unidimensional in nature, indicating that all areas of the program are considered an integral part of the whole experience. The survey instrument is appended. (SW)