A control factor for studies of high school peer influence on college aspiration was used. Previous estimates of high school peer influence on college aspirations have used peer similarity as an indicator of peer influence but have neglected to control for peers' initial similarity in aspirations at the beginning of their friendships. Longitudinal sociometric data on choices and aspirations were used to control for friends' initial similarity, and a correction factor for peer influence was computed. Two standard types of models were used: the "Wisconsin"-type recursive model of status attainment and a nonrecursive "reciprocal influence" model. The Wisconsin-type model used data from a 15-year followup study for 728 boys who chose new best friends between fall 1957 and spring 1958. The recursive-influence model followed 993 boys and 936 girls who had new best friends. Additional variables include: I.Q. and friend's I.Q.; grade point average; socioeconomic status; occupational aspiration, and educational aspiration of the child and the friend; parents' educational and occupational encouragement; and educational and occupational attainment. It was found that prior estimates of peer influence had been inflated by over 100 percent. High school peer influence on college aspirations had a weak effect. (SW)