This study presents an evolutionary process of secularization that integrates a theoretical model, simulations, and an empirical estimation that employs data from 32 countries (included in the International Social Survey Program: Religion II – ISSP, 1998). Following Bisin and Verdier (2000, 2001a), it is assumed that cultural/social norms are transmitted from one generation to the next one via two venues: (i) direct socialization – across generations, by parents; and (ii) oblique socialization – within generations, by the community and cultural environment. This paper focuses on the transmission of religious norms and in particular on the 'religious taste for children'. The theoretical framework describes the setting and the process leading to secularization of the population; the simulations give more insight into the process; and 'secularization regressions' estimate the effects of the various explanatory variables on secularization (that is measured by rare mass-attendance and by rare-prayer), lending support to corollaries derived from the theory and simulations. The main conclusions/findings are that (i) direct religious socialization efforts of one generation have a negative effect on secularization within the next generation; (ii) oblique socialization by the community has a parabolic effect on secularization; and (iii) the two types of socialization are complements in 'producing' religiosity of the next generation.