The many studies comparing good and poor readers have yielded few conclusive findings as a result of a number of problems. The first set of problems has to do with the choice of tests or tasks. There are many differences in the cognitive demands of reading and listening tasks children encounter in school. In listening, prosodic cues facilitate sentence analysis, and teachers provide external aids in retrieving relevant prior knowledge, focusing attention on main ideas, and monitoring comprehension. Determining main ideas and monitoring attention are especially important in taking advantage of the permanence of written language. As psychologists and educators turn their attention to reading comprehension, rather than individual word recognition and decoding, the differences between listening and reading warrant careful study. It is likely that causes of many reading problems will be found in the skills necessary in reading that differ from the skills children have mastered in listening. Another set of problems has to do with subject sampling and experimental design and measurement. Good discussions of these problems have been available for many years, but inadequate procedures continue to appear in published studies, and results from these studies continue to be accepted in review articles. Even when the procedures are adequate, there are serious problems in interpreting the results of studies comparing good and poor readers. The studies do not identify causal factors of reading disability nor do they recognize the great diversity of patterns of reading disability within each group of students. (HOD)