A new HR system was introduced in a Dutch hospital. The system implied collaborative work among its users. The project planning seemed to be reasonably straightforward: the system’s introduction was intended to take place gradually, including pilots in different departments and appropriate feedback. After some time, the system was successfully adopted by one group of users, but failed with another. We conceptualize the implementation process of groupware as group learning to frame the adoption of the system, and analyze the qualitative data collected during the longitudinal case study. We found that in the user group with strong group learning, adoption of the system occurred effectively and on time. In another user group with rather weak group learning, the use of the system was blocked after a short time. The results provided a first confirmation of our assumption about the importance of group learning processes in the implementation of groupware.