Teachers' judgments of student achievement are a crucial aspect of instruction and intervention. Previous research examining teachers' judgments has primarily focused on oral reading fluency and results are mixed with respect to how accurate teachers are. The purpose of the current study was to examine the accuracy of kindergarten and first-grade teachers' judgments of students' instructional level in early reading and early math and, using a pre-post quasi-experimental design, determine the extent to which the use of data altered teachers' judgments. Teachers were least accurate when rating students whose skills were in the frustrational range and they tended to overestimate students' skills. The use of data did not improve accuracy overall, but there were slight (albeit not significant) changes at the frustrational and instructional levels. Results suggest that teachers' low accuracy, when compared with screening data, combined with overestimation, may result in students who most need intervention being missed.