In a recent contribution to the Journal of Conflict Resolution, Monica Toft presents new data and analysis that suggest a rise in religious civil war, and question our findings on linguistic and religious cleavages effect on armed conflict onset. We welcome Toft’s contribution, in particular her replication of our results, and the very well documented new dataset on civil wars. Yet we disagree with several of Toft’s specific claims and raise three larger questions about the study of religion and armed conflict. First, what is the relevant counterfactual when analyzing the religion-conflict link? Second, how useful is it to classify entire conflicts as religious? Third, what is the mechanism that links religion and conflict? Revisiting her analysis and adding new empirical evidence, we show that Toft fails to undermine our earlier results and does not find support for her own theoretical mechanism of religious outbidding in the Muslim world.