Under the influence of a primary prevention trial of coronary heart disease carried out for 10 years (1971-1982) in male subjects aged 40 to 60 years at entry, the curves of the seasonal variation of the incidence of myocardial infarction, sudden death and stroke were changed. The most important reductions of the incidence of these events were registered in spring and in September. In the rest of the year the effect of the action becomes variable according to the month and to the particular major events. These observations are discussed as related to their possible cause and to the necessity of a change of action methodology according to the period of investigation.