The Graduate Record Examinations (GRE) Cooperative Validity Studies Project began in 1975 to fill the need for cooperation between graduate schools and testing agencies and for graduate level validity studies. More than 150 data sets from the 39 participating schools, representing over 19 fields of study, were analyzed. The data confirmed earlier studies which supported the correlational validity for GRE Aptitude and Advanced tests and undergraduate grade point average (GPA) with the grade point averages of first-time, full-time graduate students. Findings indicated recurring problems which must be solved in graduate-level validity studies: small samples, unstable weighting, range restriction, and criterion selection. However, it was determined that basic validity studies can be conducted, despite these problems. Limited evidence was also provided for the validity of GRE tests for females, minorities, and foreign students. It was concluded that recurring participation in these kinds of studies is more important in graduate programs with small departmental samples as units of analysis than in programs with large entering classes. (Selected findings, survey instruments, and a list of participating schools are appended.) (CP)