We study whether and how payment procedure affects behavior in a repeated public good game. To this end, we conducted five treatments, varying the times of paying the initial endowment (week before, before, or after the decisions) and the source of the initial endowment (windfall vs. earned money in a real-effort task). We find little evidence for effects of tangible endowment, windfall endowment, or prepaid endowment. However, we observe that payment procedures do influence decisions in a repeated public good game experiment.