Before and after 10 dogs were near-drowned with fresh water, cardiac filling pressures were measured during spontaneous respiration, controlled mechanical ventilation with ambient expiratory airway pressure, continuous positive-pressure ventilation (CPPV) with 20 ml H2O PEEP, and CPPV alone. Pulmonary arterial occlusion and left ventricular end diastolic pressures were measured and compared. Intrapleural pressure was subtracted from values for each of these pressures to calculate respective transmural filling pressures. Mechanical ventilation and CPPV each decreased thoracic venous return, but only CPPV increased pulmonary arteriolar resistance. The increase of both airway pressure and pulmonary arteriolar resistance, in turn, increased both right atrial and pulmonary arterial occlusion pressures, but decreased left ventricular filling. Thus, measurement of pulmonary arterial occlusion pressure alone did not allow accurate assessment of cardiac filling pressure. The authors found that measurement of intrapleural pressure was necessary to obtain an accurate reflection of left ventricular filling pressure during CPPV. Momentary interruption of CPPV to measure any pressure was of no value in assessing vascular filling and caused pulmonary edema in several animals. Therefore, the authors recommend that vascular pressures be measured and evaluated without interruption of positive airway pressure.