1. Laboratory studies have shown that cold exposure causes an increase in blood pressure, cholesterol and erythrocyte count. However, whether the mild cold exposures received during everyday life are sufficient to cause such changes is unclear. 2. To test this, outdoor temperatures in central London between 1986 and 1992 were related to both haematological and blood pressure data on 50-69-year-old men attending BUPA health screening examinations in London, and to mortality in South-East England. Since any association with temperature may be an artifact due to common, temperature-independent, annual rhythms in the parameters, these data were also analysed after removal of these circannual components by digital filtering. 3. It was found that short-term falls in temperature produced significant increases in Hb, erythrocyte count, packed cell volume, mean corpuscular Hb concentration, serum albumin, systolic and diastolic blood pressure, and significant decreases in mean corpuscular volume and erythrocyte sedimentation rate. Mean corpuscular Hb, leucocyte count, platelet count and serum cholesterol concentrations were unchanged. Time-series analysis showed that these changes occurred almost immediately in response to a fall in temperature, but persisted for longer intervals of up to 1-2 days. 4. Mortalities from ischaemic heart disease and cerebrovascular disease were also significantly increased by short-term falls in temperature. 5. These finding indicate that in the general population the cold exposures of normal life are sufficient to induce significant and prolonged haemoconcentration and hypertension, which may explain why deaths from arterial disease are more prevalent in the winter.