The purpose of this study was to examine theories of reading comprehension and instruction as they relate to instructional practices in the middle grades. Following an extensive review of literature in the theoretical areas upon which reading programs are normally based--reading, language, learning, and teaching--specific characteristics within each area were identified, and a program rating instrument was constructed. This instrument was used by three teams of specialists to assess seven reading programs in classroom situations. The results indicated that the programs do not have consistent or coherent theoretical bases; what differences they do have are obscured by their internal inconsistencies and their large areas of overlap. Even though the program analyses did not yield any significant differences, four major theoretical positions were identified that might be useful in future discussions of reading programs. (A bibliography and six appendixes of data are attached.) (RL)