Diabetes mellitus and its complications are an important cause of mortality in Western populations. The purpose of the present study was to examine the relationship between self-reported diabetes mellitus, gender, attained level of education, and socio-economic resources to all-cause mortality risk in a simple random sample of 39055 subjects, aged 25 to 74 years. Follow-up data were obtained for a maximum of 16 years, from baseline (1979-1985) to 31 December 1995. Diabetic males (2.2% of the male study group) had a relative risk (RR) for total mortality of 2.24 (CI = 1.96-2.57), adjusted for age, education, marital status, housing tenure, and car ownership, compared with non-diabetic males. The corresponding figure for females with diabetes (1.9%) was RR = 3.67 (CI = 3.16-4.27). Diabetic women had the highest age-adjusted mortality risk for coronary heart disease (CHD) of 8 compared with non-diabetic women. The corresponding RR for men was just below 3 (p<0.0001). Males and females (with and without diabetes) of low attained educational level had a RR = 1.26 (CI = 1.15-1.39) and RR = 1.54 (CI = 1.31-1.81), respectively. When analysing all people with diabetes separately, adjusting for sex and age, low-educated subjects had a 40% excess all-cause mortality compared with high-educated subjects. We conclude that diabetic women have a very high relative risk for CHD mortality compared to non-diabetic women. Furthermore, diabetic people with a low attained level of education, have an increased vulnerability to, and a higher total mortality.