This study used a combination of qualitative and quantitative research designs to analyze and evaluate procedures used in selecting youth for state-supported residential high schools for mathematics and science. Pre-admission and post-admission data were gathered for 742 students in 7 schools. Predictor variables included home school grade point average, standardized aptitude test scores, interview ratings, file ratings, and composite scores. Results indicated that students' home school adjusted grade point average was the best predictor of first and second year grade point averages. The Scholastic Aptitude Test was the second best predictor. Overall, statistical prediction was superior to professional prediction by interview or rating of complete files. Analysis of enrollment data found that African Americans and Hispanic students are proportionally underrepresented, while Asian students are proportionally overrepresented. Administrators saw the use of multiple criteria as a major strength of their identification systems, and viewed the lack of minority representation as a major weakness as was the relatively high rate of attrition. Appendices list the schools and the administrator interview questions. (Contains approximately 65 references.) (DB)