Mortality and morbidity in elderly people are higher in winter than in summer months, with seasonal variations in rates of both fatal and non-fatal myocardial infarction and stroke. To identify factors that might contribute to the excess winter frequency of cardiovascular disease in the elderly, we studied 100 subjects aged 75 and over who lived in either their own homes or in sheltered or residential accommodation. Each person was visited each month for one year, body and environmental temperatures were noted, and cardiovascular risk factors were measured. 32 subjects withdrew from the study. Significant seasonal effects were found for fibrinogen, plasma viscosity, and HDL cholesterol (p less than 0.003, Bonferroni adjustment). Plasma fibrinogen concentrations showed the greatest seasonal change and were 23% higher in the coldest six months compared with summer months. Fibrinogen was significantly (p less than 0.05) and negatively related to core body temperature and all measures of environmental temperature. Those living in institutions had greater changes in plasma fibrinogen than those living in the community. The seasonal variation in plasma fibrinogen concentration is large enough to increase the risk of both myocardial infarction and stroke in winter.