During 1981-82 approximately 17 percent of all Graduate Records Examinations (GRE) General Test takers were foreign nationals from over 140 countries. Heavily concentrated in scientific-technical fields, foreign nationals made up almost one-half of all engineering and one-third of all math-science examinees. Most were non-native English speakers whose average verbal and analytical scores were considerably lower than those of native English speakers (both U.S. and foreign) in similar fields. Foreign examinees' English language background did not affect the quantitative mean profile which was similar to that of U.S. examinees. Detailed comparative profiles of GRE performance data are provided for the general foreign and U.S. examinee populations, and for foreign examinees classified by country of citizenship. In addition, the study analyzes GRE group performance of foreign and U.S. examinees by age, sex, reported English language communication status, intended graduate field, U.S. vs other undergraduate origin, repeater vs non-repeater status, and other personal, academic, and testing-related characteristics. Study findings suggest that GRE General Test data (especially verbal and analytical test data) generated by non-U.S. citizens from non-native English speaking societies should not be treated as comparable to test data for U.S. examinees or examinees from other native-English speaking societies.