Hypercholesterolemia or hypertension are continuous risk factors for coronary heart disease. When a preventive action is carried out against such a risk factor, it is necessary to specify a risk factor level value, named the treatment threshold, above which a subject should be treated. But a non-arbitrary determination of this threshold value is impossible from the epidemiological data. A method for the non-arbitrary determination of the treatment threshold value is presented, based upon mathematical modelling of the clinical and economics consequences of the prevention policy in the whole population. In a cost-effectiveness approach, the model was used to estimate the cost per coronary event prevented according to the threshold value for blood cholesterol. It was found that a minimum in this outcome proposed as the optimum treatment threshold. It is possible, from a public health point of view, to determine an optimum, non-arbitrary treatment threshold value in the prevention of coronary heart disease with cholesterol-lowering interventions. The model proposed here could be useful in decision making processes.