This paper reports the findings of a study that explored the validity of curriculum-based measures of reading in order to develop a valid and practical instrument for regular education teachers as well as special education teachers in Korea. Participants in the study were 213 students from 7 elementary schools in Seoul, Korea. Of the total, 183 second-grade students were placed in the regular classrooms, 30 students (grades 1-6) with Individualized Education Programs (IEP) for reading were also from the special classes. Testing materials consisted of four different tasks: oral reading tests, question-answering tests, a maze procedure, and a standardized reading comprehension subtest. In general, the results revealed that the reliabilities of the reading inventories were adequate. The scores of the oral reading tasks showed greater correlations with the criterion measures than those of the rest of the reading inventories. In addition, the oral reading measures produced a substantial discrimination (regular education students/students with reading IEPs). The subsequent analyses also revealed that the oral reading measures provided much higher validity coefficients for the students with reading IEPs. Charts illustrate statistical results. (Contains 24 references.) (Author/CR)