One of the personality characteristics which may be assumed to influence group behavior is the tendency for members to play ascendant or submissive roles. It might be supposed that the greater the tendency toward ascendance, the more likely a member would be to engage in leadership acts, with the extent to which he actually serves as a leader being determined by the degree to which ascendance tendencies are present in other members of the group. In order to rest hypotheses about the postulated relationship of ascendance to member behavior and group position, it is important to know whether valid measures of ascendance can be readily obtained. There are several personality inventories which purport to measure ascendance or dominance and their reliabilities are in general fairly satisfactory. Data on the validity of these scales, however, is relatively meager. Accordingly, one aim of a study of the effects of varying combinations of ascendant and submissive group members was to check the validity of the 30-item ascendance scale of the Guilford-Zimmerman Temperament Survey (1). The Survey was administered to approximately 350 introductory psychology students. Male Ss who received scores above 20 and below 12 were then used to form four $-person groups of each of the following types: 4 submissive Ss, 4 ascendant Ss, 1 submissive and 3 ascendant Ss, and 1 ascendant and 3 submissive Ss. When the groups met, one member of each was asked to serve as the leader in order to facilitate the carrying out of two group tasks. A member was picked at random from the first two types of groups, while the one submissive S and the one ascendant S were picked from the third and fourth types, respectively. Two hidden 0s independently recorded group behavior and also classified each S as to whether they felt his behavior was ascendant or submissive. 0s were never told the composition of a group, nor were they present when a member was designated to serve as the leader. The inter-observer reliability for a random sample of groups was .86. For the purpose of reporting the validity of the ascendance scale, only the analysis of appointed-follower activity will be given here since being designated the leader appeared to reduce activity differences between ascendant and submissive Ss. Analyses of variance showed submissive followers to be significantly less active than ascendant followers, regardless of the type of leader (P = 8.84,~ = < .02). Ascendant followers asked significantly more questions of each other, gave more opinions and suggestions, and acted upon and rejected more suggestions than did submissive followers. 0s' agreement and disagreement with scale classification of ascendance-submission were 69% and 31%, respectively, a significant departure from chance expectancies (2 = 6.75, p= <.01). REFERENCE 1. GUILFORD, J. P., & ZIMMERMAN, W. S. The Guilford-Zimmerman Temperament Survey. Beverly Hills, Calif.: Sheridan Supply Co., 1949.