Substantial empirical evidence suggests that alcohol consumption is an important cause of cirrhosis mortality levels. However, recent research has failed to find the expected positive relationship between consumption and cirrhosis mortality, in both the United States and Canada, when data from the mid-1970s through the late 1980s are analysed. Although recent studies have investigated a number of possible explanations, this study examines the hypothesis that an increase in the provision of treatment for alcoholism resulted in a disjuncture in the established relationship between consumption and cirrhosis deaths. This hypothesis is evaluated with a multivariate time series model in which the relationship between consumption and cirrhosis mortality is estimated with controls for treatment and the unemployment rate, using data from the US state of North Carolina. The results indicate that with treatment controlled, changes in cirrhosis mortality are independent of consumption. Further, treatment has a significant short-term lagged effect on cirrhosis mortality, suggesting that the impact of treatment on chronic alcohol abusers may be one of delaying the consequences of such abuse.