Two studies were carried out to examine socio-economic factors in breast cancer: a random sample of all new cases in Edinburgh in 1979 was reviewed, and the control population of the Edinburgh randomized trial of breast screening was used to determine stage and survival in relation to social class. Small area statistics from census data were used as measures of social class, the method being now well accepted. More than one-third of women still present with obviously advanced or metastatic breast cancer, but both studies showed this has no association with socio-economic status. Late stage at presentation is a serious problem, and although mass screening is likely to cause an improvement in those who are screened, it cannot in those who do not attend for screening. As attendance is related to social class, less affluent women are less likely to benefit and will continue to be diagnosed with advanced disease.