The enforcement of social norms often requires that unaffected third parties sanction offenders. Given the renewed interest of economists in norms, the literature on third party punishment is surprisingly thin, however. In this paper, we report on the results of an experiment designed to evaluate two distinct explanations for this phenomenon, indignation and group reciprocity. We find evidence in favor of both, with the caveat that the incidence of indignation-driven sanctions is perhaps smaller than earlier studies have hinted. Furthermore, our results suggest that second parties use sanctions to promote conformism while third parties intervene primarily to promote efficiency.