During a study of group communication, efforts were made to identify variables that distinguish statements of consensus groups from those of non-consensus groups, and to determine if and how relationships between consecutive statements differ for the 2 types of groups. Six students in each of 3 consensus and 3 non-consensus groups discussed the University of Iowa's policies on grading, undergraduate women's hours, and undergraduates' possession of automobiles. Fifty pairs of consecutive statements were selected from each discussion and two-factor analyses of variance were made on the first statements in each pair on 8 separate variables. The variables were clarity of expression, opinionatedness, amount of information, interest, provocativeness, orientation, objectivity, and length of contribution. No differences were found between the 2 types of groups on either clarity or length. The consensus group's statements revealed much more interest in women's hours and much less interest in undergraduates' possession of cars, and the group's discussion on this question was less opinionated, and more provocative, informative, and objective than those of the non-consensus group. The study has shown that there are some detectable differences in the discussion behavior of groups that reach consensus and groups which do not. A list of 8 conclusions is included, and statements on substantive, methodological, and educational implications for future research. (WM)