Three models predict an association between urbanism and nontraditional behavior: (1) that it is a function of the characteristics of individuals found in cities; (2) that it is due to the anomie of cities; (3) that it is due to the generation of and consequent influence of innovative urban subcultures. Secondary analysis of American survey data on religiosity, church attendance, attitudes toward alcohol and birth control confirm the general urbanism-deviance association. Controlling for covariates suggests that Model 1 is inadequate for there remains an independent effect of residence—albeit a quite small one—unaccounted for by individual traits. Some suggestive data point to Model 3 as the more accurate one.