A study analyzed the validity of inferential, cause/effect, and main idea questions which were asked in five selected commercial informal reading inventories (IRIs). The inventories were "Analytical Reading Inventory (3rd Edition),""Basic Reading Inventory (4th Edition)"; "Burns and Roe Informal Reading Inventory (3rd Edition)"; "Classroom Reading Inventory (5th Edition)"; and "New Sucher-Allred Reading Placement Inventory." Passages and questions for grades 1 through 6 on Forms A and B from each IRI were examined. Results indicated that: (1) many questions labeled as main idea were actually asking for the topic of the reading passage; (2) many IRIs contain narrative passages which contain neither a stated nor an implied main idea; (3) inference questions were often ambiguously worded, asked opinions from the reader, or were not inferential in nature; (4) many inference questions failed to ask students to explain the reasoning behind their answers to the question; (5) only two of the IRIs included cause/effect questions, but these questions were labeled correctly by the authors of the tests more often than the other types of questions studied; and (6) for some tests, the different forms were inconsistent in the number of questions and the manner in which they were asked. Findings suggest that a more open-ended questioning or retelling format would allow a more accurate evaluation of comprehension of a specific passage. (Two tables of data are included.) (RS)