The authors asked 74 teachers to predict average- and low-performing students' reading performance using a rating scale and actual curriculum-based measures of oral reading fluency. The authors compared teachers' judgments with students' performance on curriculum-based measures and a standardized achievement measure of reading. Comparisons between the accuracy of teacher judgment, as reflected in correlational relations among measures, and a how-close analysis showed that although correlations between predicted teacher scores and actual student performance ranged from moderate to strong, teachers did poorly in a how-close analysis comparing actual with predicted student scores. Teachers had lower levels of accuracy for lower achieving students. The authors discuss practical implications relative to teacher judgments of student performance. (Contains 4 tables.)