Experienced teachers enrolled in two graduate reading classes examined seven informal reading inventories (IRIs)--three at the elementary level and four at the secondary level--to (1) discover what text types (narrative or expository) they used at each level to measure student comprehension skills and determine instructional levels and (2) identify the rhetorical structures used in expository passages. The 18 teachers who rated the elementary passages and the 20 who rated the secondary ones had all previously administered IRIs and had studied the literature on text type and structure and reading comprehension. Each teacher examined at least four IRIs, classified each passage used in them as narrative or expository, judged each narrative passage as well- or poorly-formed, and judged each expository passage according to rhetorical structures adapted from B. J. F. Meyer (1975). The teachers found that all of the preprimer and primer passages used on the IRIs were narrative, while most of the other elementary passages were narrative and most of the secondary passages expository. In addition, they found that many of the narrative passages on the IRIs were not well-formed, and that approximately one-eighth of the elementary and one-fourth of the secondary expository passages had no clear rhetorical structure. The findings suggest that the passages used on IRIs might produce erratic comprehension scores. In light of these findings, five practical suggestions are offered for teachers and diagnosticians who use the currently available commercial IRIs. (FL)