Hemodynamic studies were performed before and after pericardiocentesis in 19 patients with pericardial effusion. Right atrial pressure decreases significantly, from 16 +/- 4 mm Hg (mean +/- SD) to 7 +/- 5 mm Hg in 14 patients with cardiac tamponade. This change was accompanied by significant increases in cardiac output (3.87 +/- 1.77 to 7 +/- 2.2 l/min) and inspiratory systemic arterial pulse pressure (45 +/- 29 to 81 +/- 23 mm Hg). The remaining five patients did not demonstrate cardiac tamponade, as evidenced by lack of significant change in these hemodynamic parameters. In all patients with tamponade, right ventricular end-diastolic pressure (RVEDP) was elevated and equal to pericardial pressure; equilibration was uniformly absent in patients without tamponade. During gradual fluid withdrawal in the tamponade group, significant hemodynamic improvement was largely confined to the period when right ventricular filling pressure remained equilibrated with pericardial pressure. In 10 patients with tamponade and pulsus paradoxus, pulmonary arterial wedge pressure (PAW) was equal to pericardial pressure except during early inspiration and expiration when it was transiently less and greater, respectively; however, inspiratory right atrial pressure never fell below pericardial pressure. In these 10 patients, PAW decreased significantly following pericardiocentesis (P less than 0.001). In the remaining four patients with tamponade but without pulsus paradoxus, all of whom had chronic renal failure, PAW was consistently higher than pericardial pressure or RVEDP and did not decrease after pericardiocentesis. These data tend to confirm the hypothesis that in patients with tamponade, the venous pressure required to maintain any given cardiac volume is determined by pericardial rather than ventricular compliance. When pericardial compliance determines diastolic pressure in both ventricles, relative filling of the ventricles will be competitive and determined by their respective venous pressures (pulmonary vs systemic), which vary with respiration and alternately favor right and left ventricular filling. This results in pulsus paradoxus. However, if pulmonary arterial wedge pressure is markedly elevated before the onset of tamponade, as in patients with chronic renal failure, then pericardial compliance may only determine right ventricular filling pressure. In such cases, pulsus paradoxus may be absent.