To investigate the pathophysiology of cardiac tamponade, the hemodynamics of 77 consecutive patients with greater than 150 ml of pericardial effusion were studied. Patients were classified into 3 groups based on the equilibration of intrapericardial with right atrial and pulmonary arterial wedge pressures (mm Hg): group I (n = 16), intrapericardial pressure was less than right atrial and pulmonary arterial wedge pressures; group II (n = 13), intrapericardial pressure was equilibrated with right atrial but not pulmonary arterial wedge pressures; group III (n = 48), intrapericardial pressure was equilibrated with right atrial and pulmonary arterial wedge pressures. Pericardiocentesis produced the following changes: group I--significant (p less than 0.03) decreases in intrapericardial pressure (7 +/- 2 mm Hg), right atrial pressure (3 +/- 2 mm Hg), pulmonary arterial wedge pressure (2 +/- 2 mm Hg), and the inspiratory decrease in arterial systolic pressure (3 +/- 4 mm Hg) but no significant change in cardiac output; group II--significant (p less than 0.02) decreases in intrapericardial pressure (11 +/- 5 mm Hg), right atrial pressure (6 +/- 4 mm Hg), pulmonary arterial wedge pressure (4 +/- 5 mm Hg), and inspiratory decrease in arterial systolic pressure (8 +/- 7 mm Hg), and increase in cardiac output (1.1 +/- 1.2 liters/min); group III--significant (p less than 0.001) decreases in intrapericardial pressure (16 +/- 7 mm Hg), right atrial pressure (9 +/- 4 mm Hg), pulmonary arterial wedge pressure (8 +/- 5 mm Hg), inspiratory decrease in arterial systolic pressure (17 +/- 11 mm Hg), and increase in cardiac output (2.8 +/- 1.5 liters/min). The changes after pericardiocentesis in all parameters were significantly (p less than 0.05) greater in group III than in groups I or II except for the change in right atrial pressure, which was not significantly different in groups II versus III. The changes after pericardiocentesis indicate pericardial effusion caused the greatest abnormalities in group III but also caused significant abnormalities of pressure and flow in group II and of pressure alone in group I.(ABSTRACT TRUNCATED AT 250 WORDS)