Previous research has shown that individuals who are more religious use alcohol less than those who are less religious. The purpose of the current study was to investigate the relationship between various dimensions of religion and alcohol use, including an examination of the potential mediating role of social cognitive constructs in that relationship. It was hypothesized that: 1) Individuals who were more religious would drink alcohol less often, consume fewer drinks per drinking day, and experience fewer alcohol-related problems than those who were less religious. 2) Individuals who utilized more self-regulatory strategies, had higher self-efficacy, had less positive and more negative expectancies of alcohol use, and had standards of more moderate alcohol use would report less frequent alcohol use, fewer drinks per drinking day, and fewer alcohol-related problems. 3) Individuals who reported greater religiosity would utilize drinking self-regulatory strategies more frequently, have higher self-efficacy for avoiding heavy alcohol use, have less positive and more negative outcome expectancies regarding alcohol use, and have standards of more moderate alcohol use. 4) The relationships between the dimensions of religiosity and the alcohol use indices would be mediated by the social cognitive constructs.

Three hundred and thirty-four college students completed a number of self-report questionnaires assessing alcohol use, religiosity dimensions, and social cognitive constructs. Results were consistent with, and expanded upon, previous research by demonstrating that various dimensions of religiosity were inversely related to, and also multivariate predictors of, the quantity and frequency of alcohol use and the experience of alcohol-related problems in a sample of college students. Findings further suggested that this relationship was completely mediated by several social cognitive constructs including self-efficacy, outcome expectancies, internal standards regarding alcohol use, and the use of self-regulatory strategies. Limitations of the study and suggestions for future research are discussed.