Two experiments investigated the effects of sources of strategy information on children's acquisition and transfer of reading outcomes and strategy use. Children with reading skill deficiencies received comprehension instruction on main ideas. In the first experiment, the final sample comprised 33 students (21 fourth graders, 12 fifth gaders) drawn from one elementary school. The 19 boys and 14 girls ranged in age from 9 years 7 months to 12 years 7 months. Although different socioeconomic backgrounds were represented, children predominantly were lower-middle class. Ethnic composition of the sample was: 40% Hispanic American, 28% Black, 26% White, 6% Asian American. Teachers initially nominated 34 children for participation; one student was randomly excluded from the appropriate cell to equalize condition sizes. Subjects regularly received remedial reading comprehension instruction. Students had been placed in remedial classes by the school district because they scored at or below the 30th percentile on the reading subtest of the SRA (Science Research Associates) Survey of Basic Skills, 1985. Some students were taught a comprehension strategy, while others received strategy instruction and strategy value feedback linking strategy use with improved performance, and controls received comprehension instruction without the strategy. In the second experiment, students (N = 33, 13 boys, 20 girls) were drawn from one elementary school. Ages of the 15 fourth graders and 18 fifth graders ranged from 9 years 9 months to 12 years 4 months. All students were enrolled in remedial reading classes because they scored in the lowest 30th percentile of the reading subtest of the SRA Survey of Basic Skills. Subject characteristics and selection procedures were similar to those of Experiment 1. Ethnic composition of the sample was: 46% Hispanic American, 30% White, 18% Black, and 6% Asian American. These students were taught the comprehension strategy or received instruction without strategy training, after which they were given comprehension instruction on locating details. Some children were taught to modify the strategy; others did not employ the strategy on details. Results indicated that children who received strategy value feedback (in the first experiment) and strategy modification instruction (in the second experiment) demonstrated the highest self-efficacy, skill, strategy use, and transfer. Findings support the idea that remedial readers benefit from information about strategy usefulness. (Two tables of data are included; 45 references are attached.) (RS)