A study was undertaken to (1) examine the development and construction of a Group Informal Reading Inventory to predict the reading comprehension levels (independent, instructional, and frustration) of junior high school bilingual students for the purpose of reading instruction; and (2) validate the inventory through a three-way correlational study comparing the comprehension results with those of a cloze test, a standardized test, and a questionnaire by which teachers estimate students' reading levels. The study involved 50 bilingual students of predominantly English- and Spanish-speaking, low- and middle-income backgrounds in an urban school. All had been instructed in Spanish until they gained English language proficiency; then they were mainstreamed into the English curriculum. It was discovered that the students were all functioning far below their developmental grade levels and their assigned present grade levels, and native language grades were lower than those in English. It is recommended that (1) a decision be made for each individual student as to whether he should be taught in two languages or, if his native language skills are insufficient to transfer to English as a second language, whether he should be taught in English; (2) testing for reading and content areas be administered regularly to monitor progress; and (3) there be careful regulation of the timing, techniques, content, materials, and evaluation of bilingual instruction. (MSE)