BACKGROUND: The use of immunosuppressive therapy for myocarditis is controversial. PURPOSE: To review the literature on the effectiveness of immunosuppressive therapy in biopsy-proven lymphocytic myocarditis. DATA SOURCES: Two authors independently searched MEDLINE and other medical databases from 1980 to 26 June 1997. STUDY SELECTION: Randomized, controlled trials; matched-cohort studies; and case-control studies of patients with biopsy-proven myocarditis (Dallas criteria or a mean of > or = 2.5 lymphocytes per high-power field) for which any form of immunosuppressive treatment was used. The outcomes of interest were mortality and change in left ventricular ejection fraction. DATA EXTRACTION: 6 of 374 studies satisfied the selection criteria. DATA SYNTHESIS: In survivors, left ventricular function in myocarditis improved approximately 10% over 6 months without immunosuppressive treatment. Prednisone alone did not improve survival (P >0.2) or left ventricular function (P >0.11). Prednisone combined with azathioprine or cyclosporine did not improve survival (P >0.2) or left ventricular function (P >0.2) in three studies. However, one small matched-cohort study showed improvement in children (P <0.01). Neither interferon nor thymic hormone improved survival or left ventricular function. CONCLUSIONS: Immunosuppressive therapy is ineffective in lymphocytic myocarditis. Current therapy in lymphocytic myocarditis seems to be limited to supportive measures or transplantation.