Some of the problems with the traditional measures of socioeconomic status include (1) the loss of information resulting from combining different factors that have varying associations with health problems; (2) the reverse causal pathway that exists from health and illness to income and occupation; and (3) a number of particular problems with deriving socioeconomic status from census tract information. In contrast there are clear advantages to using educational status as the primary socioeconomic index. A wide variety of literature is reviewed pointing to a strong positive relationship between years of schooling and health. Three models that attempt to account for this association are described. It is suggested that the educational status of patients should be part of their data base.