The inotropic effects of albumin were studied in 94 seriously injured patients who received an average of 14.5 transfusions, 9.2 liters of crystalloid and 0.9 liters of plasma prior to end of operation; 46 patients, by random selection, received added albumin averaging 31 gm during operation, 198 gm during the early postoperative period of extravascular fluid sequestration, and 395 gm during the first 4 days of the later fluid mobilization period. Left ventricular stroke work index (LVSWI) was plotted against pulmonary wedge pressure (Ppw) in 22 patients who had indwelling thermistor pulmonary artery catheters at the time of the first study. Calculated heart work units (WU) were derived from the pulse pressure, mean arterial pressure, pulse rate, and central venous pressure (CVP) in patients without LVSWI measurements. Albumin supplementation increased serum albumin (4.2 vs. 2.9 gm%), plasma volume, CVP (15 vs. 9 cm H2O), but did not alter red cell volume (1,531 vs. 1,519 ml). The ratio of LVSWI/Ppw fell in albumin patients (1.9 +/- 1.6 vs. 4.8 +/- 1.8), and the ratio of WU/CVP was significantly depressed in albumin patients (4.9 +/- 2.3 vs. 7.3 +/- 2.1). The slopes of the LVSWI/Ppw and WU/CVP were shifted to the right in albumin patients. This negative inotropic effect was associated with impaired oxygenation, as reflected by an increased ratio of inspired oxygen per arterial oxygen tension (0.62 +/- 0.06 vs. 0.33 +/- 0.1). Finally, 24 of the 46 albumin-treated patients were digitalized for heart failure, compared to only 11 of the 48 nonalbumin patients. Pending subsequent studies, albumin should be considered a potentially negative inotropic agent.