The research reported in this paper tested two major claims made by Emile Durkheim more than one hundred years ago: first, that Protestants are more likely to commit suicide than Catholics, and second, that this greater vulnerability of Protestants to self-destruction is due to their lower level of social integration. Of these two statements our results confirmed the first but not the second: while Protestants have preserved - despite the profound historic changes that took place during the last several decades - their increased susceptibility to suicide, the explanation Durkheim proposed was inconsistent with many of our findings. First, contrary to what Durkheim's ideas would lead us to expect, including church attendance as a control variable did not make denominational differences disappear; in fact, it made them even stronger. And second, we found interaction effects that were hard to reconcile with Durkheim's theory: while this theory would predict religious differences to decline with increasing attachment to the church community, what actually happened was just the reverse: as the degree of social integration, measured by church attendance, increased, the gap between Protestants and Catholics widened.Therefore, it seems that in order to explain the impact of social integration on suicide, or on deviant behavior in general, we should not focus exclusively on the role it plays in reducing loneliness or individualism but also on the role it plays in conveying specific norms, values and behavior patterns. This would not only facilitate a more complete understanding of how social integration works, but would also help connect two distinct branches of theories of deviance: control theory and subculture theory.