A n experiment explored possible personality differences between sociometric techniques following small group discussions. Such differences could predict emergence of leadership behavior in initially leaderless discussion groups. Participating in assigned groups of four or five were 269 Master of Business Administration students and 65 business executives. Discussion sessions lasted a maximum of 30 minutes. Students and research assistants observed the groups and assessed specified behavior. A sociometric questionnaire was used after the session, group members ranking each other on amount of participation, guidance of discussion, best ideas, leadership, and popularity. A six hour battery of psychological tests was given. Statistical analysis indicated a significant relationship of personality, interest, attitude, and the amount of participation to sociometric choices. The shortness of the discussions may have decreased reliability and predictability of behavior. (ja)