Clinical factors of possible importance for the greater than two-fold rise in the incidence of Kaposi sarcoma of the elderly in Sweden before the AIDS epidemic were reviewed in 63 regional patients. 5 patients had lymphoproliferative disease before or at the time of Kaposi sarcoma, and 4 patients had been receiving steroids (including 1 with lymphoma) at diagnosis. 2 of these 9 patients plus 2 additional patients had received blood transfusions 1-9 years before diagnosis. None of 17 patients tested was positive for HIV-1, and none had signs of an unexplained progressive immune defect. Of the evaluable cases, 27% had diabetes mellitus and 7% had had previous myocardial infarction. However, only the frequency of congestive heart failure (47%) was significantly greater than that of an ambulatory control group (P = 0.001) in the age group 75-84 years. Exposure to cytomegalovirus (CMV) was not more common in 15 Kaposi sarcoma patients than in an age and sex matched control group. No single factor could account for increased Kaposi sarcoma among the elderly. If the classical form has an infectious aetiology, the tumour could arise after effective transmission of the agent (as by a transfusion), especially combined with some degree of immune deficiency or perhaps congestive failure late in life.