To clarify the controversy of endomyocardial biopsy (EMB) in terms of its diagnostic value, we performed a meta-analysis of EMB studies published between 1982 and 1993, including our own experience. A total of 255 articles was retrieved using both a computer search of the Medline database and a manual bibliographic search, but only 30 studies with 4,313 patients met the predefined inclusion/exclusion criteria. The diagnostic value of EMB was classified into four categories, according to the effect of EMB findings on the discharge diagnosis: aetiology uncovered, new diagnosis of heart muscle disease (HMD) revealed, clinical diagnosis confirmed, and no useful information obtained. Clarification of aetiology of HMD was reported in 28 out of 30 studies with a total of 4,195 patients and it was achieved by EMB in 17.9% of these patients (95% confidence interval (CI) was 16.8-19.1%). A new unexpected diagnosis of HMD was arrived at in 25 of 30 studies (3,947 patients) and this occurred in 19.3% of patients (95% CI = 18.1-20.6%). Confirmed clinical diagnosis of HMD by EMB was covered by 12 studies (1,231 patients) and was proven in 40.1% of patients (95% CI = 37.3-42.7%). EMB not providing any useful clinical information was mentioned in seven of 30 studies (857 patients); this happened in 5.9% of patients (95% CI = 4.5-7.4%). Therefore, these results confirmed the remarkable diagnostic value of EMB. It was equally helpful in all diagnostic categories and had considerable overall diagnostic utility.