Most experimental group research (whether conducted in an Electronic Meeting System (EMS) environment or not) has used ad hoc groups (i.e., randomly assembled groups of individuals that have no history and no future) rather than established groups that have a history and probable future of working together. This may be one factor that partially accounts for the differences in research findings between prior EMS experimental and field research. The present study, which examined otherwise similar ad hoc and established groups working in similar EMS environments, found that, while the total amount of communication within both types of groups was similar, there were significant differences in the type of communication that occurred. Established groups made more uninhibited comments and more process-oriented comments, had a less equal distribution of participation among the group members, and perceived a lesser focus on the task. While there were similar levels of direct conflict within the two types of groups, members of ad hoc groups were more likely to 'sugar-coat negative comments. For ad hoc groups, conflict was negatively correlated with the quality of meeting outcomes. Perhaps the most interesting finding for future research is that there was a significantly higher degree of variability among the established groups across virtually all measures, i.e., ad hoc groups tended to be more similar, while established groups exhibited distinct personalities.