The results of different types of reading achievement measures were compared for 58 low-income urban black third graders. Two formal tests were administered: the norm-referenced California Achievement Tests (CAT), and the criterion-referenced Alabama Basic Competency Test (ABCT). Informal measures included the Houghton-Mifflin Informal Reading Inventory (HMIRI), the Classroom Reading Inventory (CRI), a cloze procedure, and teacher judgment (as indicated by the basal reader assignment for each student). Results indicated that correlations among all of the measures were moderate to high. The formal tests, particularly the CAT, tended to produce lower scores than the informal measures. In spite of high correlations, the CAT and ABCT results revealed very different distributions of student ability. With the ABCT, more students showed average and above average performance. Examination of teachers' judgments regarding reading book placement, as compared to test results, indicated that teachers underestimated students' reading ability and placements did not reflect test results. HMIRI results also suggested that a number of students could have been assigned to a higher-level reading book. It was suggested that informal measures be used for book placement and that multiple measures of reading achievement be used in decision making. (GDC)