This study investigated how students collaborate in a CSCL environment and how this collaboration affects group performance. To answer these questions, the collaborative process of 101 groups of secondary education students when working on a historical inquiry task was analyzed. Our analyses show that group members devote most of their efforts to regulation of task-related activities. For example, by formulating plans or strategies or monitoring task progress. Group members also engaged in social activities often (e.g., disclosing personal information, joking). Less attention was paid to exchange of task-related information (e.g., asking task-related questions) and regulation of social activities (e.g., planning and monitoring the collaboration). Exploratory factor analysis was used to identify the interrelationships between the different collaborative activities. This analysis showed that collaborative activities can be grouped in four broad categories: discussion of information, regulation of task-related activities, regulation of social activities, and social activities. These activities were then used to predict group performance using multiple regression analysis. No effect of discussion of information and regulation of task-related activities on group performance were found. Regulation of social activities positively affected group performance, whereas social interaction negatively affected group performance. As in this study no inferences could be made about the causal relation between collaboration and performance, future research should attempt to focus on this relationship, for example by investigating more closely how different individual and group factors affect collaboration and group performance. (Contains 8 tables.)