The purpose of this study was to examine the technical (psychometric) adequacy of office discipline referrals (ODRs) as a behavioral assessment tool for individual student, data-based decision making within a problem-solving model. Participants were 990,908 students in 2,509 elementary schools. Students were grouped into ODR cut points of 0-1, 2-5, and 6 or more total ODRs. Descriptive analyses indicated consistent mean ODR growth trajectories for each group. Logistic regression analyses showed that intermediate cut points (1 ODR, 2 or more ODRs) were moderately accurate in predicting total ODRs received, and accuracy increased throughout the fall. Specific problem behaviors were less predictive of total ODRs, although adding specific behaviors to intermediate cut points enhanced prediction to some extent. Results are discussed in terms of using ODRs to identify students in need of intervention and monitor progress. (Contains 4 tables, 1 figure and 1 footnote.)