BACKGROUND AND OBJECTIVE: Air temperature changes are associated with increased cardiovascular and respiratory risk, but the roles of inflammatory and coagulation markers are not well understood. We investigated the associations between temperature and several blood markers in patients with coronary heart disease (CHD) and pulmonary disease (PD). METHODS: Two studies were conducted in Erfurt, Germany, over two successive winters. 578 and 381 repeated blood measurements were collected from 57 CHD and 38 PD patients, respectively. Data on patient characteristics and disease history were gathered at baseline. Meteorological data were collected from existing networks. Associations were analysed using additive mixed models with random patient effects. Effect modification by diabetes status was investigated only in CHD patients, as only two PD patients had diabetes. RESULTS: Mean daily air temperature varied between -13 degrees C and 16 degrees C in both study periods. A 10 degrees C decrease in the 5-day temperature average before blood withdrawal led to an increase in platelet counts (% change from the mean: 3.0%, 95% CI 0.6% to 5.5%) and fibrinogen (5.5%, 1.3% to 9.7%), no change in C-reactive protein in PD patients, and a decrease in C-reactive protein in CHD patients. A 2-day delayed increase in factor VII associated with temperature decrease was seen in CHD patients (4.9%; 0.7% to 9.2%), while PD patients showed no effect. 'Effects in CHD patients without diabetes' into 'Effects on factor VII in CHD patients without diabetes'. CONCLUSIONS: This study suggests that temperature decrease is associated with change in several blood parameters. The complex interplay of blood markers at low temperature may contribute to the observed association between cold and cardiovascular mortality and morbidity.