The study examines how level of reading acquisition is related to performance on word boundary tasks written at four differing readability levels. It also explores the validity of using word boundary tasks as measures of reading ability by comparing student performance on a word boundary measure to performance on both a cloze procedure task and on a word recognition task. Level of reading acquisition, level of passage readability, and grade placement of students were found to affect performance on the word boundary task. A multitrait-multimethod matrix of intercorrelations between the three measures showed that the word boundary task used for this study appeared to be a reliable and valid measure of reading performance for these subjects and that the three methods measure some of the same traits.Grade level scores established by standardized reading tests are often used by teachers to determine the level at which reading instruction is provided. Unfortunately, this determination does not ensure successful reading of instructional materials since standardized test grade-equivalent scores do not always provide an accurate estimate of the student's instructional level (Harris and Sipay, 1975;Sipay, 1964). This problem seems compounded in the content areas by the large range of levels within texts, the above grade readability level of many texts (Johnson, 1975) and the teachers' uncertain estimates of the level of reading difficulty of the textbooks (Olson, 1970).Since Betts' (1946) book, Foundations of Reading Instruction, was first published, some teachers have favored informal measures which assess individual reading ability. The cloze procedure (Taylor, 1953) has been utilized with some success for this purpose, and there is some empirical support for its use as a measure of comprehension (Bormuth, 1969;Rankin, 1959) and as a measure of the readability of passages (Bormuth, 1968; Taylor, 1953).