This article uses a Canadian national sample to examine the relationship between conventional religious belief, church attendance, and belief in paranormal phenomena. Greater religious belief is strongly associated with greater paranormal belief. Church attendance (and other measures of religious participation) are only weakly associated with paranormal belief until conventional religious belief is statistically controlled; once this is done, greater church attendance is strongly associated with lowered paranormal belief. Together, these two religious variables explain about one-quarter of the variance in paranormal belief, making them the strongest predictors that have yet to be identified.