Eight-hundred thirty patients (pts) with suspected myocardial disease of undefined etiology were observed from 1978 to 1996. In 350 pts, the clinical diagnosis was of dilated cardiomyopathy (DCM) or myocarditis. An endomyocardial biopsy was performed on all patients and in 54 of them (15%), an active myocarditis was identified. In six cases, myocarditis was detected at autopsy. There were 37 male patients and 23 females, with an average age of 35.5 +/- 15 years (range 1.67). Mean time interval between clinical onset and diagnosis was 4 +/- 10 months. Clinical presentation was characterized in 4 cases by fulminant myocarditis (Group I), in 8 cases by chest pain (Group II), in 14 cases by arrhythmia (Group III: hypokinetic in 9 pts and hyperkinetic in 5) and, in the last 34 pts, by congestive heart failure (CHF) (Group IV). Improvement was defined at 9 +/- 3 months according to a clinical score based on left ventricular shortening fraction (increase > or = 5 units), New York Heart Association Class improvement by (at least one Class) and left ventricular end-diastolic diameter (decrease > or = 10%). The main clinical and instrumental parameters characterizing the groups were: a more severe dilatation and left ventricular dysfunction in the pts belonging to Group I or IV with respect to those in Group II and III; a significantly worse prognosis in terms of evolution in DCM or death/cardiac transplantation (CT) in the pts from the Group II and III. After a follow-up period of 48 +/- 46 months, the mortality in the four groups was: 100% (4/4), 0% (0/8), 21% (3/14), 38% (13/34). Fifty percent of deaths were concentrated in the first 2 years of follow-up. Left ventricular end-diastolic diameter (OR 1.09, p < 0.05), age (OR 0.95), presence of left ventricular bundle branch block (OR 2.32), right ventricular function (OR 2.43) at clinical onset and the status of improvement at 9 +/- 3 months of follow-up (OR 0.24, p < 0.05) are predictors of evolution in DCM or death/CT for the pts with onset from CHF (Group IV). Immunosuppressive treatment has been utilized for the 76% of the pts. No conclusion can be drawn on the efficacy of this therapy, but no adverse events significantly related to therapy have been observed in a 9 +/- 3 months follow-up period. In conclusion, myocarditis can show a clinical presentation polymorphism, which influences the prognosis and natural history of the disease. Evolution in DCM and adverse events (death/CT) are more common in Groups I and IV. Some simple parameters evaluated at clinical presentation and the proposed classification as "improved" or "not improved" after a short-term follow-up (9 +/- 3 months) show good predictive accuracy. The present study does not allow us to draw any conclusion about the efficacy of immunosuppressive treatment. A randomized, controlled, large-scale trial, with adequate follow-up and advanced histological diagnosis techniques will help define the role of immunosuppressive therapy and patient eligibility criteria for this treatment.