The survival data on 515 white and 388 black female breast cancer patients seen at the Medial College of Virginia between 1968 and 1977 were analyzed to study the effect of age, stage and race on survival prognosis. For a subset of the data representing patients from the city of Richmond (117 white and 206 black), socio-economic status (SES) information was generated on the basis of six predictors of SES and, in addition, the role of social class was studied. Each of these factors has a significant association with survival time. In particular, the probability of surviving a given length of time after diagnosis is ordered according to the socio-economic level and the statistical test for dose response show a highly significant directional relationship. Age and stage do not explain the difference in survival between the two races. Race and SES are highly associated; a higher proportion of blacks than whites come from the lower end of the socio-economic scale. Moreover, the racial difference in survival becomes insignificant when it is adjusted for the distribution of socio-economic levels. This suggests that the observed difference in breast cancer survival between blacks and whites is, to a large extent, due to the difference between the two races with respect to the distribution of socio-economic status.