This article describes a conceptual model developed to explain abusive use of alcohol and reports results of an initial test of that conceptualization. General social support is hypothesized to affect the level of subjective well-being, while alcohol-relevant social support affects the degree of alcohol involvement. A cross-sectional test of two models derived from this formulation was made using data from 148 alcoholic clients entering treatment at a private non-profit psychiatric facility. Results indicate that a model in which no direct relationship between alcohol involvement and subjective well-being is specified provides a more parsimonious explanation of interrelationships at the time of treatment entry. Alcohol involvement is explained by alcohol-relevant affiliative and instrumental support (albeit weakly), and subjective well-being is explained by general affiliative and instrumental social support. There is virtually no interrelationship between alcohol involvement and subjective well-being once the support variables are taken into account. The findings indicate that treatment should more often incorporate others from a client's social network, using significant others first to provide general social support and later alcohol-relevant support as well. Further analyses involving samples with more heterogeneity in levels of social investment and exploring the utility of these variables for treatment matching categorization decisions are warranted.