Health promotion activities have come under attack because of experimental evidence that reduction of cardiovascular risk does not appreciably affect total mortality. Reviewing the results, McCormick and Skrabanek are sceptical about health promotion in general. Their arguments embody the assumption, often shared by proponents, that the purpose of health promotion is life extension. In our opinion, the primary purpose of most health promotion activities in developed societies is to improve quality of life, to “compress” morbidity, and to extend active life expectancy. The compression of morbidity is illustrated in the figure. Extension of life itself can be a realistic goal as long as the aim is to prevent deaths that occur reasonably early—for example, by seat-belt laws and preventive measures in infancy. There also remain opportunities to lessen mortality in developing societies and in disadvantaged populations where improvements of social conditions and personal risk factors have lagged behind those in more favoured groups. But, for health promotion in general, we must face some new facts.The disturbing new factsLife expectancyLooking at the most recent and best performed trials of primary prevention of cardiovascular disease—the MRFIT Study, the Lipid Research Clinics Study, the Physician's Aspirin Study, and the Helsinki Heart Study—we agree with McCormick and Skrabanek that there is no effect whatsoever upon total mortality (table I). This lack of effect is unlikely to be due to a counterbalancing mortality from side-effects of drugs since it is seen also in trials of diet or exercise.Epidemiological data support these observations. It is commonly stated that if atherosclerosis, the cause of 49% of all deaths in the United States, were eliminated, the average life expectancy would rise 8-10 years. However, in Japan there is essentially no atherosclerosis, and the Japanese national average serum cholesterol is very low. The average Japanese does live a little longer than the average American, but the difference in life expectancy at advanced ages is only a few months, not 8 or 10 years. (Higher incidence of other diseases in Japan, such as gastric carcinoma and stroke, does not begin to account for this observation.) Similarly, natural experiments on good health practices have been underway in the United States over many years. In Utah, because of the pervasive presence of the Mormon Church, smoking and drinking rates are very far below national averages.