The present study aimed to test the effect of role socialization processes on declines in drinking after marriage. Role socialization as it relates to marriage theorizes that after occupying a role, individuals are likely to change their behaviors to conform to role expectations of marriage, such as reductions in drinking (Yamaguchi & Kandel, 1985). Previous literature has examined declines in drinking behaviors after marriage and inferred that role socialization was the underlying process. This study is the first to directly test whether beliefs that alcohol is harmful to the marital role predicted declines in frequency of drunkenness after marriage. Ordered probit regression was used to test the effect of marriage-related motives to limit drinking on declines in frequency of drunkenness from before marriage to after marriage. Analyses revealed that marriage-related motives to limit drinking were not significantly predictive of declines in frequency of drunkenness after marriage. Only partner drinking emerged as a significant predictor of declines in frequency of drunkenness after marriage. These results highlight the need for a reliable and valid measure of role socialization processes as they relate to the marital role. Furthermore, future studies should consider studying participants at different time points after marriage and consider measuring commitment to the marital role as a moderator. Such studies will help to better understand the results of this study as well as better understand the marriage effect on drinking.