This study examined the predictive contribution of occupational and marital stress to the development of symptomatolgoy, affective disorder, and alcohol-related problems in a sample of 325 male power plant employees interviewed at baseline and one-year follow up. After controlling for known clinical and social risk factors, we found statistically significant predictive effects of job demands on affective disorder and job decision latitude on alcohol problems. In addition, the specific combination of high job demands with low decision latitude was important in predicting the occurrence of alcohol problems. Marital stress was not predictive of poorer mental health. The analyses provided only minimal support for a stress-buffering rote of social support.