To examine the relationship between degree of treatment-induced blood pressure (BP) reduction and myocardial infarction, the experience of 1765 previously untreated, mild to moderate hypertensives (initial BP greater than or equal to 160 and/or greater than or equal to 95 mm Hg) in a systematic treatment program was reviewed. Over an average of 4.2 years, there were 39 morbid or fatal myocardial infarctions. Three types of fall in diastolic BP (final minus pretreatment) were defined: small (less than or equal to 6 mm Hg), moderate (7 to 17 mm Hg), and large (greater than or equal to 18 mm Hg). By Cox regression, an association was observed between myocardial infarctions and both a large and a small fall, relative to a moderate decline. Age and sex were independent risk factors. Body mass index, cholesterol level, electrocardiogram, race, prior cardiovascular disease, smoking status, initial and final revisit BP, total intreatment BP, and systolic BP were not. Since both a large and small reduction in diastolic BP were associated with a higher incidence of myocardial infarction (relative to a moderate fall), perhaps a moderate reduction in diastolic BP should be the goal of treatment for mild and moderate hypertensives.