The role of hemostatic variables (which promote hemostatic plugs and thrombi) and rheological variables (which affect blood flow) in the pathogenesis of vascular diseases (ischemic heart disease, stroke, and peripheral arterial disease) is reviewed, with emphasis on epidemiological studies. Rheological variables are consistently associated with both prevalent and incident cardiovascular disease. These associations are only partly explained by conventional risk factors. The predictive value of plasma viscosity for cardiovascular events is partly explained by fibrinogen, and partly by lipoproteins. The associations of whole blood viscosity with cardiovascular disease are partly explained by plasma viscosity and partly by hematocrit. White cell count, but not platelet count, predicts ischemic heart disease events. Cigarette smokers have higher levels of rheological variables than non-smokers, these increases are partly or wholly reversible in ex-smokers. Lipoprotein reduction by pravastatin lowers plasma and whole-blood viscosity, which may be one mechanism through which lipid lowering produces an early reduction in cardiovascular events. Data from the Edinburgh Artery Study suggest that viscosity is related both to the extent of atherosclerosis, and to ischemia in the presence of a given degree of atherosclerotic stenoses. Among hemostatic variables, fibrinogen, factor VIII: vWF complex, tpA antigen, and fibrin D-dimer are associated with both prevalent and incident cardiovascular disease. Again, these associations are only partly explained by conventional risk factors They suggest that endothelial disturbance and increased fibrin turnover may play roles in cardiovascular disease. Hemostatic and rheological variables are therefore associated with both prevalent and incident cardiovascular disease, and may be mechanisms through which risk factors such as smoking, hyperlipidemia and infections (including oral infections) promote vascular events.