This study addresses three key issues related to media differences among computer-mediated groups meeting face-to-face, synchronously, and asynchronously: First, do groups using different media perceive characteristics of these media differently? Second, do media-related perceptions remain static or change as they are used over time? And finally, do media differences result in performance differences, i.e., do richer media result in better performance for equivocal tasks as predicted by bandwidth theories? Some results of this study tend to support media-characteristics theories while others offer limited support for more evolutionary perspectives. For instance, initially, face-to-face groups found their medium to be warmer, have a better interface, and be more effective compared to their distributed counterparts. While many of these initial perceptions lingered over time, there was one notable exception. At the start of the study, face-to-face groups rated their medium as being more effective than synchronous groups; however, by the end of the study, no significant differences were apparent. Moreover, despite the persistently lower social presence of leaner media, distributed-synchronous groups performed better than their face-to-face counterparts. Finally, the two types of distributed groups--synchronous and asynchronous--did not differ significantly in their perceptions or performance. This study improves our understanding of distributed interaction while simultaneously highlighting the need to further investigate the relationships among tasks, technologies and teams over time.