Distinguishing constrictive pericarditis from restrictive cardiomyopathy is a difficult clinical challenge. We review published reports in which hemodynamic criteria were used to differentiate these two diagnoses. There were 82 cases of constriction and 37 cases of restriction. The overall predictive accuracy of the difference between right and left ventricular end-diastolic pressures (RVEDP and LVEDP), RV systolic pressure, and the ratio of RVEDP to RV systolic pressure were 85%, 70%, and 76%, respectively. If all three criteria were concordant, the probability of having correctly classified the patient was greater than 90%. However, one fourth of patients could not be classified by hemodynamic criteria. There are few data to support the use of hemodynamic measurements after exercise or volume infusion to separate these two groups. Numerous recent studies have reported on the ability of left ventriculography, Doppler echocardiography, or radionuclide angiography to distinguish constriction from restriction. Many of the proposed indices appear promising, but these studies suffer from small sample size, potential selection bias, and complexity of the proposed criteria, which have limited their widespread application. New imaging technologies, such as CT scanning or MRI have been applied in a limited number of cases, but appear to be a sensitive means of detecting abnormal pericardium. Endomyocardial biopsy has proven useful in establishing the diagnosis of infiltrative cardiomyopathies, eliminating in those cases the need for surgical intervention. The finding of myocarditis must be considered a nonspecific finding that does not preclude thoracotomy. Since constrictive pericarditis is a surgically curable condition, the distinction between constrictive and restrictive disease is of critical importance. Taking into account the relative contribution of data derived from hemodynamic, imaging,and biopsy studies, we propose an algorithm for the selection of appropriate candidates for pericardial biopsy and stripping.