Many digitally mediated peer-production systems allow participants to define their own activities. The challenge in such systems, however, lies in retaining members beyond the first few interactions. To address this problem we must understand who these users are and why they begin to contribute. Importantly, there is scant empirical evidence on how motivations are associated with different trajectories of participation for new participants. Our study addresses this gap by combining a survey of new Wikipedia editors' motivations with an exploratory analysis of the editors' activity logs. Using clustering techniques to identify prototypical activity profiles from log data, we observe what motivations are associated with which prototypical activities. We find that new editors' motivations are predictive of their future activity. In particular our results indicate that reputation, social, enjoyment, and obligation motives differ among editor activity clusters.