Public school teachers (N = 189) role-played as members of school councils making principal selection decisions by rating simulated candidates for principal vacancies. The independent variables were principal candidate job experience, candidate person characteristics, and teacher school level. The dependent variable was teacher rating of the job candidate. A three-way ANOVA detected a significant main effect (F [2, 162] = 7.34, p less than 0.001) for candidate job experience. Teachers rated the most experienced candidates higher than the least experienced candidates, but failed to rate the medium experienced candidate higher than the least experienced candidate, or the most experienced candidate higher than the medium experienced candidate. Implications are discussed relative to the practice of teachers selecting principals, selection theory, and future research. (Contains 3 tables.)