Since 1975, the National Association of Secondary School Principals (NASSP) has attempted to improve the selection process for all levels of principals through the establishment of regional assessment centers. Presently, 37 centers operated by districts, consortia, universities, and state departments of education are producing data being used by approximately 4,000 school districts. The center being studied is university-operated, produces data for nine school districts, and is called the Center for the Assessment of Administrative skills (CAAS). This paper examines the internal validity of the process for selecting school principals at CAAS. NASSP supplied both the materials and the trained observers to be assessors. The study specifically examines (1) the variables of uniqueness and influence on overall ratings; (2) assessor reliability; (3) agreement of assessor ratings for each participant; and (4) persistent variance in participant performance for various demographic groupings. Results confirm that the eight centers conducted by CAAS have operated well within the standards required by NASSP, the sanctioning body, and within the broader standards required for effectively assessing candidates' potential. Further research is recommended to study assessor bias. Included are 17 references and 9 tables. (MLH)