Comparisons were made of test scores derived from a sample of American Indian students attending grades 2 through 10, living in a Bureau of Indian Affairs dormitory, and enrolled either in an English as a Second Language Program (N=43) or in the regular school program (N=87) to determine ESL impact on Indian students' English vocabulary, comprehension, and scholastic ability. Measurements were made via use of: the Wide Range Achievement Test (WRAT, administered as pre- and post-tests to both groups); the Gates-MacGinitie Reading Test (given to 9th and 10th grade students only); and Grade Point Average (GPA, used to determine who went into the ESL program). The study period was five months. Analyses of covariance were employed for purposes of comparison and hypotheses testing. Results indicated no difference in reading level advancement due to ESL treatment. Although there was confirming evidence for reliability (test-retest) and concurrent validity (correlation) for the WRAT and the Gates-MacGinitie standardized tests, the relatively consistent negative relationships between these measures and grades proved puzzling. It was concluded that close examination of teacher grading criteria and standarized test objectives would provide some insight into the discrepancy. (JC)