Participation is one of the main conditions for successful computer supported self-directed learning of groups (Carell et al. 2005). In contrast, research concerning participation in online courses or discussion forums show clearly, that many participants act more or less passiv or even break off (Lipponen et al. 2002). This high rate of free-riding could have multiple reasons (Cress 2005). To find out the underlining reasons most of the related research had focused on an individual level of participation. But with respect to group learning, it is also important to look at group participation and who the group members participate and act together. Accordingly, studies increasingly look into participation in online courses in a more sophisticated manner. With respect to group learning, these approaches focusses on who and on which levels online-communication occurs and who it can be can be enhanced to intensify group learning. Hitherto, there are no studies who investigate the quality of participation with respect of self-directed learning of groups. Hence, the aim of this study was, to describe and analyze patterns of participation in an online course with respect to self-directed learning. In the self-directed learning context we defined participation quantitatively as both a) total time students spent in the learning environment, b) as the process of participation, and qualitatively as c) the roles or functions the students played in their group. An empirical field study was held as a course at the University. It lasted 64 days. The course was accessible by students studying educational sciences. 20 students divided into three groups participated in the course. Data was gathered by log-file analysis, interaction and network analysis of chat-protocols and from students' self-reports in online-interviews after the end of the course. An empirical field study was held as a course at the University. It lasted 64 days. The course was accessible by students studying educational sciences. 20 students divided into three groups participated in the course. Data was gathered by log-file analysis, interaction and network analysis of chat-protocols and from students' self-reports in online-interviews after the end of the course.The main results were: 1) Compared with traditional face-to-face courses, participation in the online course was both higher and more regular. 2) Hierarchical cluster-analysis showed four patterns of participation, which differed in the total amount of time spent in the learning-environment and in the process of participation. 3) There was a high congruence between patterns of participation and group-membership. 4) In each group one or two persons adopted implicitly the role of group process moderator. These persons were also those most engaged at all levels of interaction (coordination, content, socio-emotional). 5) There was a strong relationship between the engagement of students in deep processing learning on one hand, and both the total and process of participation on the other. But in contrast, those who showed the highest levels of participation and whose participation was very stable over the whole course, were found only to be engaged in surface level learning. 6) There is evidence that in particular those students stopped their participation who are highly interested in deep processing learning and effective learning respectively.Cress, U. (2005). Why member portraits can undermine participation. In: Koschmann, Tim; Suthers, Dan; Chan, T.W. (Eds.): Proceedings of CSCL 2005. The next 10 Years. Mahwah, New Jersey: LEA. Carell, A. et al. (2005): Improving the Coordination of Collaborative Learning with Process Models. In: Koschmann, Tim; Suthers, Dan; Chan, T.W. (Eds.): Proceedings of CSCL 2005. The next 10 Years. Mahwah, New Jersey: LEA, 18-27. Kollock, P. & Smith, M. (1996). Managing the Virtual Commons: Cooperation and Conflict in Computer Communities. In S. Herring (Hrsg.), Computer-Mediated Communication: Linguistic, Social, and Cross-Cultural Perspectives (109-128). Amsterdam: John Benjamins.Lipponen, L. et al. (2002). Effective Participation and discourse through a computer network: Investigating elementary students' computer-supported interaction. Journal of Educational Computing Research, 27(4), 355-384. Wellman, B. & Gulia, M. (1999). Net surfers don't ride alone: Virtual communities as communities. In B. Wellman (Ed.), Networks in the Global Village (167-194). Boulder: Westview Press.