The main purposes of the Northwick Park Heart Study (NPHS) are to improve the prediction of ischaemic heart disease (IHD) and to elucidate its pathogenesis. Measures of haemostatic function are made along with those of variables already known to be associated with IHD; one such variable is cigarette smoking. It is clearly of interest to know whether the smoking effect may be mediated through changes in haemostatic function; data at recruitment in 1392 men and 580 women have therefore been examined in order to study such changes. Since certain characteristics in NPHS, particularly age and social class, are correlated with both cigarette smoking and with levels of some haemostatic variables, multiple regression analysis has been used to identify smoking effects independent of associations between smoking and these other characteristics. The haemostatic variables included in this, analysis are factors V, VII and VIII, fibrinogen, antithrombin III, fibrinolytic activity, platelet count and platelet adhesiveness. In men, factor VIII levels and fibrinolytic activity are significantly lower in smokers than non-smokers; fibrinogen levels are significantly higher. In women, factor VIII levels are also significantly lower in smokers, but there is no difference in fibrinogen; for fibrinolytic activity there is an interaction between smoking and the use of oral contraceptives (OC). Activity in non-smokers on OC is substantially higher than in those not on OC, but activity in smokers on OC is the same as in women not on OC. It therefore seems that smoking abolishes the protective rise in fibrinolytic activity caused by OC. In men, cigarette smoking may promote thrombosis through effects on fibrinogen and fibrinolytic activity.