Volume loading is used to treat hemodynamically compromised patients with acute pulmonary embolism despite data to suggest that volume loading after embolism might cause a leftward shift of the ventricular septum with a subsequent decrease in left ventricular (LV) end-diastolic volume and stroke work. We studied 10 closed-chest, anesthetized, and ventilated dogs to assess the effects of volume loading after pulmonary embolism caused by autologous clot. LV, right ventricular, and right atrial pressures as well as LV anteroposterior, septum-to-right ventricular, and septum-to-LV free wall diameters (sonomicrometry) were measured. Pericardial pressure was measured with flat, liquid-containing balloons. The effects of volume loading were assessed before embolism, after one episode of embolization, and after repeated embolizations. The LV area index (as a reflection of LV volume) increased during volume loading before embolism (2,870 +/- 430 to 3,080 +/- 520 mm2; p less than 0.05), did not change significantly during infusion of fluid after one embolization (2,850 +/- 470 to 2,860 +/- 500 mm2; p = NS), and decreased significantly during volume expansion after repeated embolizations (2,760 +/- 440 to 2,660 +/- 420 mm2; p less than 0.01). An index of LV stroke work increased (188 +/- 85 to 260 +/- 101 mm Hg x mm2; p less than 0.05), did not change significantly (188 +/- 39 to 203 +/- 52 mm Hg x mm2; p = NS), and decreased markedly (133 +/- 64 to 45 +/- 27 mm Hg x mm2; p less than 0.001) before embolism, after one embolization, and after repeated embolizations, respectively. The decrease in LV area index during volume loading after repeated embolizations was associated with an increase in septum-to-right ventricular free wall diameter (31 +/- 8 to 34 +/- 8 mm; p = 0.001) and a decrease in the septum-to-LV free wall diameter (44 +/- 5 to 42 +/- 5 mm; p less than 0.001), whereas the LV anteroposterior diameter did not change (62 +/- 5 to 63 +/- 5 mm; p = NS). This is compatible with a leftward septal shift being partially responsible for the decrease in LV end-diastolic volume; such a shift would be expected with the observed decrease in transseptal end-diastolic pressure gradient (-3 +/- 2 to -5 +/- 2 mm Hg; p = 0.001). In addition, after repeated embolizations, LV transmural pressure decreased in response to the volume load reflecting a marked increase in pericardial pressure.(ABSTRACT TRUNCATED AT 400 WORDS)