OBJECTIVE The aim of the present study was to investigate whether the nitric oxide donor sodium nitroprusside can reduce the cardiac inflammatory response during coronary artery bypass grafting in patients with severely compromised left ventricular function. METHODS Patients (n = 30) were assigned to receive placebo or sodium nitroprusside (0.5 microg. kg(-1). min(-1)) for the first 60 minutes of reperfusion. Interleukin 6, interleukin 8, and tumor necrosis factor alpha levels; platelet adhesion molecule CD41 and CD62 levels; and CD11b on leukocytes were determined in the radial artery and coronary sinus before cardiopulmonary bypass and during reperfusion (1, 5, 10, 35, and 75 minutes). RESULTS At 1 minute of reperfusion, coronary venous levels of CD41-positive polymorphonuclear leukocytes were 8% lower than arterial levels in the placebo group and 18% higher in the sodium nitroprusside group (P =.021). At 5 minutes of reperfusion, the respective levels were 29% and 1% for interleukin 6 (P =.015), -5% and 20% for CD41-positive monocytes (P =.032), and -2% and 16% for CD11b-positive monocytes (P =.038). At 10 minutes of reperfusion, these levels were -14% and 21% for CD41-positive monocytes (P =.006). At 35 minutes of reperfusion, these levels were -13% and 7% for CD41-positive monocytes (P =.017), -41% and 23% for CD11b-positive monocytes (P =.001), and 7% and 25% for CD62-positive platelets (P =. 041). At 75 minutes of reperfusion, the levels were 15% and -7% for tumor necrosis factor alpha (P =.025) and -10% and 10% for CD62-positive platelets (P =.041). CONCLUSIONS Transcardiac production of proinflammatory cytokines is reduced in patients undergoing coronary artery bypass grafting treated with the nitric oxide donor sodium nitroprusside. At the same time, less activated leukocytes and platelets are retained in the coronary circulation.