This study analyzes the relationship between religions and economic development by using country-level panel data. This research contributes to the literature by examining the economic effect of religious denomination and the nonlinear relationship, and by using split sample regressions. The regression analysis shows that i) Islam has a negative effect on the level of GDP, but is positively related to GDP growth, although the regression results are not robust; ii) there exists an inverted U-shaped relationship between Protestantism and GDP level; and iii) the economic effects of religion are different across the levels of income. Overall, the regression analysis shows a substantial contrast between Western religions and non-Western religions. The empirical results suggest that the economic effect of religion depends on the peculiar characteristics of each religion and the mediating role of the level of economic development.