When people opt out of participating in one sphere of civic life, are they more likely to opt out of others? Is civic disengagement a diffuse or a domain-specific process? This study addresses these questions by re-evaluating researchers' assumptions about the relationship between religious engagement and political participation. Using validated measures of voter turnout, I examine whether the growing population of religiously-unaffiliated Americans is more or less likely to vote than their affiliated counterparts. Results demonstrate that the "secular voting gap" is smaller than previous estimates would indicate, and they challenge conventional wisdom about the positive relationship between attendance at religious services and increased political participation. These results support domain-specific theories of disengagement that are sensitive to the substantive meaning of different religious practices. *Reading the working paper in advance is required for attendance.* Email cura@bu.edu for your copy. Co-sponsored with the School of Theology