A 3-year project developed a procedure to achieve the objectives of using miscue analysis in the reading program in a way that is economical in time, applicable to more than one student at once, and usable as a teaching tool as well as a diagnostic tool. Three separate trials, in 1975, 1976, and 1977, refined and developed the procedure until it could be used with groups of up to eight students at a time and produce reading improvement. Students were involved in analysis of the tapes after one student had read each day. All eight students were invited to add to the tell-back, and to take part in the discussion of the miscues. Students reported benefits from the procedures even before they had an opportunity to read. Each of them read at least once and took part in the analysis of the other readings. The importance of the students' self-image, and of the tutor's support for positive self-image, did not form part of the original hypothesis but emerged during the trials as a vital aspect of the whole program, inseparable from reading improvement. Conclusions emerging from the three years of project work establish the viability of this procedure as a useful teaching tool, while leaving other questions to be answered in follow-up projects. (Two figures are included and 13 references are attached. Appendixes include transcripts of interviews held with all participants in the 1976 projects and case studies from the 1976 project.) (MG)