We examined 10 common indicators of social isolation and assessed which, if any, are related to subjective well-being among elderly individuals. The quantitative measures (no daily contact and minimal weekly contact), no children, having no children plus being unmarried, and having no children plus living alone, were unrelated to either global happiness or life satisfaction. The combined indicators of living alone and being unmarried as well as single indicators of having no companions or having no confidants were related at the bivariate level to both measures of well-being. When controlling for demographic, economic, and health factors, having no confidants and no companions were significantly and independently related to subjective well-being. These findings suggest that practitioners who seek information on living arrangements or marital status as proxy measures of lower psychological well-being are utilizing the wrong predictors.