The results presented in this paper concerning regional obesity as a health hazard in women refer to a 12-year longitudinal population study of 1,462 women, aged 38-60, which was carried out in Gothenburg, Sweden, in 1968-69. In univariate analysis the ratio of waist-to-hip circumference showed a significant positive association with the 12-year incidence of myocardial infarction, angina pectoris, stroke, diabetes mellitus and death. The association with incidence of myocardial infarction and diabetes mellitus remained in multivariate analysis. The relation between the ratio of waist-to-hip circumference and the end points studied was stronger than for any other anthropometric variable studied.