In the Speedwell study set up to examine primarily the relation of plasma lipids and ischaemic heart disease in men aged 45 to 64 years drawn randomly from the practices of 16 general practitioners, various haemostatic factors which may contribute both to thrombogenesis and atherogenesis were measured. Fibrinogen measured nephelometrically and plasma viscosity were positively associated with the prevalence of ischaemic heart disease. Antithrombin III was negatively associated with the prevalence of ischaemic heart disease. These associations were statistically significant at the 1% level of probability. Fibrinogen measured by a clotting method and the heparin neutralising activity of platelet poor plasma had negative and positive associations with ischaemic heart disease, respectively, but neither association achieved statistical significance. Because of the interrelation of these variables and also age, a multiple logistic regression analysis was undertaken separately for the two measures of fibrinogen. The negative association of "clottable' fibrinogen and the positive association of plasma viscosity with the prevalence of ischaemic heart disease were confirmed and both were statistically significant at the 5% level of probability. Apart from age the independent association of the other variables with ischaemic heart disease did not achieve statistical significance. Fibrinogen measured by both methods had positive and statistically significant associations with serum total cholesterol, but no associations with serum total triglycerides, smoking, or alcohol consumption. "Clottable' fibrinogen had an inverse and statistically insignificant association with serum high density lipoprotein cholesterol. The observed associations support the concept of the involvement of some haemostatic factors in the aetiology of ischaemic heart disease, and these associations are now being examined more critically in a longitudinal study.