We examine the classic question of how religious diversity in a community affects church membership in a period of high growth and social change. Culling the literature, we develop hypotheses about five possible mechanisms underlying a diversity-growth relation: (1) secularization; (2) organization environment isomorphism; (3) organizational competition; (4) identity activation; and (5) capacity utilization. We test these hypotheses using panel data from 1890 to 1916, using models specified to overcome recently identified artifactual problems and estimators to deal with community-specific omitted factors. The findings support the plausibility of mechanisms based on secularization and identity activation; they contradict predictions from mechanisms based on isomorphism, competition, and capacity utilization. We also find that Protestant and Catholic religions thrived in communities that differed in nativity.