Between 1960 and 1973, a total of 104 patients at the Mayo Clinic had a diagnosis of idiopathic dilated cardiomyopathy on the basis of clinical and angiographic criteria; these patients were followed up for 6 to 20 years. Twenty-one percent of the patients had a history of excessive consumption of alcohol, 20 percent had had a severe influenza-like syndrome within 60 days before the appearance of cardiac manifestations and 8 percent had had rheumatic fever without involvement of cardiac valves several years before; thus, possible etiologic risk factors of infectious-immunologic type may be important. Eighty patients (77 percent) had an accelerated course to death, with two thirds of the deaths occurring within the first 2 years. Twenty-four patients (23 percent) survived, and 18 of them had clinical improvement and a normal or reduced heart size. Univariate analysis at the time of diagnosis revealed three factors that were highly predictive (p < 0.01) of the clinical course: age, cardiothoracic ratio on chest roentgenography and cardiac index. Systemic emboli occurred in 18 percent of the patients who did not receive anticoagulant therapy and in none of those who did; thus, anticoagulant agents should probably be prescribed unless their use is contraindicated.