This study describes intergroup guessing differences in response to tests and to test-like tasks. It is a composite of seven component inquiries with three substudies in Phase 1 and four in Phase 2. These seven studies cover the Graduate Record Examination (GRE) item-type domain from a number of viewpoints relevant to implicit guessing behavior. The studies in Phase 1 centered on item analytic strategies and on test data derived from a GRE administration. The studies in Phase 2 centered on item-component strategies and on data derived from supplementary materials administered at four university settings. The groups studied were Whites, Chicanos, and Blacks. The implication of any intergroup differences might be that the scoring formula and the instructions to candidates concerning scoring were inappropriate for one or more groups. The most general conclusion is that such intergroup differences do not exist. In seven attempts to find group contrasts, with each attempt yielding a fairly complex and multifaceted analysis, only one minor phenomenon can be reported: Chicano female omitting on GRE-Verbal is demonstrated by individual groups of somewhat lower average ability than that of individuals demonstrating similar behavior for other ethnic-sex groups. (PN)