An assessment analysis was performed to determine whether sets of items designed to measure three different subskills of reading comprehension of the Iowa Tests of Basic Skills (ITBSs) did, in fact, distinguish among these subskills. The three major skills objectives were: (1) facts; (2) generalizations; and (3) inferences. Data from administration of the ITBSs to 82 sixth-graders in the Carterville (Illinois) Grade School were analyzed. The reading comprehension test was divided for the purpose of analysis into: (1) 17 items assessing facts; (2) 22 items assessing generalizations; and (3) 17 items assessing inferences. Internal construct validity was calculated with a procedure developed by G. H. Henning (1987). Results suggest that, for this sample, the reading comprehension items did not have robust coefficients of internal construct validity. Only 17 of the 56 items exhibited internal construct validity when compared with two criterion measures. Students' strengths and weaknesses in the different comprehension skills could not be determined, because there was a lack of items demonstrating robust discrimination among these three reading comprehension skills. It is concluded that school systems mandated to show that their reading instructional programs effect mastery in different reading comprehension skills will have a difficult task of offering empirical proof if the ITBSs are used to elicit the test data. Six tables contain study data and an appendix lists skills objectives. (SLD)