This study attempted to determine the effects of a twice-weekly, four-month-long, cross-age tutoring program on children's reading achievement levels and self-concepts. A total of 240 students, stratified by grade and sex, were randomly assigned to four experimental groups. Within each group, students were matched by grade: grade four with grade one, grade five with grade two, and grade six with grade three. The experimental groups were randomly assigned to two treatment and two control situations. One treatment group worked in a special reading room, while the other worked in the library. One control group read individually in the reading room; the other did not participate in any way. An analysis of data did not prove the hypothesis in question, which asserted that the enactment of a role is influenced by both expectation and location. Research design was limited by the length of time that the individual variables operated, by the lack of base-line data, and by the use of a single school's population. It was suggested that further research be directed toward developing more specific criterion measures. (Author/KS)