In clinical cardiac tamponade, open-catheter intrapericardial pressure (IPP) may be used to estimate left ventricular transmural filling pressure (TMFP). However, it has been suggested recently that right atrial pressure (RAP) is superior to IPP in assessing true extracardiac pressure during pericardial drainage. In 10 patients with subacute cardiac tamponade, pulmonary wedge pressure (PWP), RAP, and IPP were measured along with indexes of systolic function. To test the relative merits of IPP and RAP in assessing true pericardial pressure, three TMFP estimates were analyzed: TMFP1 = (PWP - IPP); TMFP2 = (PWP - 1/3 RAP - 2/3 IPP); and TMFP3 = (PWP - RAP). An accurate TMFP presumably should increase during pericardiocentesis and correlate with left ventricular stroke work. In addition, to test the role of preload variation in pulsus paradoxus, respiratory variation in TMFP was analyzed. In the initial tamponade state, RAP and IPP were essentially equal, so all three TMFP estimates gave equivalent results. For instance, TMFP1 averaged 4 +/- 2 mm Hg but fell to 0.2 +/- 1.3 mm Hg during inspiration (p less than .001 vs expiration) and showed beat-by-beat correlation with pulse arterial pressure. After intermediate pericardiocentesis (280 +/- 160 ml), the IPP of 6 +/- 3 mm Hg fell significantly below the RAP of 10 +/- 3 mm Hg (p less than .001), but with a 570 +/- 320 ml residual effusion suggesting continued IPP measurement accuracy. By complete pericardiocentesis (810 +/- 430 ml) there was a significant increase in TMFP1 to 8 +/- 4 mm Hg (p less than .05 vs tamponade) but not in the TMFP3 of 1 +/- 3 mm Hg. Encompassing tamponade and pericardiocentesis data, left ventricular stroke work index showed positive correlation with TMFP1 (r = .59) and TMFP2 (r = .52) but not with TMFP3. Thus cardiac tamponade often may be diagnosed with a TMFP averaging well above zero, and diastolic equalization of PWP, RAP, and IPP may be a predominantly inspiratory finding ("inspiratory tracking"). This supports the role of preload variation in the genesis of pulsus paradoxus. On the other hand, true pericardial pressure may fall substantially below RAP in the course of pericardial drainage. This may be reconciled with the concept that normal pericardial pressure nearly equals RAP by hypothesizing an increased pericardial capacity in subacute tamponade so that pericardiocentesis produces a state analogous to removal of normal pericardial constraint.