Whether hierarchical logistic regression can reduce the sample size requirement for estimating optimal cutoff scores in a course placement service where predictive validity is measured by a threshold utility function is explored. Data from courses with varying class size were randomly partitioned into two halves per course. Nonhierarchical and hierarchical analyses were performed on each half. Compared to their nonhierarchical counterparts, hierarchically estimated cutoff scores from different halves were more stable and predicted course outcomes in the other half more accurately. These differences were most pronounced with small samples. Sample size requirements for developing cutoff scores for course placement can be substantially reduced if hierarchical logistic regression is used.