The fairness of testing practices for selection or promotion of members of different racial and ethnic groups has been questioned. In this study, a search was made of occupations in the Federal Government to find those in various agencies with sufficient ethnic group representation for acceptable sample size, and for which dependable and objective criterion measures might be developed. Research conducted over a six-year period focused on the occupations of Medical Technician, Cartographic Technician, and Inventory Management Specialist. After careful job analyses, aptitude and ability test batteries were selected, several types of criterion measures were developed, and background data and job activities questionnaires were prepared. Data analysis indicated that tests found to be valid for one ethnic group were in general valid for the other ethnic groups included in the study, across occupations. Regression equations developed on majority group data appeared to predict almost equally well for minority groups, and some instances predicted higher criterion scores for minorities. The use of supervisors' ratings as a criterion of job performance in validity studies with ethnic sub-groups was called into question by findings of this study, which disclosed interaction effects ascribed to ethnic group membership of rater and ethnic group membership of ratee. (Author/BJG)