Anti-CD18 monoclonal antibodies (MAb) have demonstrated variable protection against neutrophil (PMN)-mediated myocardial reperfusion injury. To identify factors contributing to this variability, open-chest dogs underwent coronary artery occlusion for 90 min followed by reperfusion for 3.5 h. Ten minutes before reperfusion the dogs received saline (n = 18) or one of three anti-CD18 MAb: MHM.23, R15.7, or PLM-2 (2, 1, and 1 mg/kg and n = 19, 8, and 4, respectively). Collateral flow was measured with radioactive microspheres, area at risk was assessed with monastral blue dye, and infarct size was measured postmortem by triphenyltetrazolium chloride. In vitro, all three MAb bound to canine PMNs, but only MHM.23 and R15.7 inhibited their adherence to keyhole limpet hemocyanin-coated plastic. In vivo, only MHM.23 and R15.7 significantly reduced infarct size after adjusting for the effect of collateral flow. MHM.23 afforded protection in dogs with moderate ischemia (epicardial collateral flow > 0.1 ml.min-1.g-1, infarct size reduced 46%) but not in dogs with more severe ischemia. Only R15.7 was effective in dogs with severe ischemia. Although MHM.23 and R15.7 produced similar inhibition of tissue PMN accumulation, as reflected by myeloperoxidase activity. R15.7 markedly inhibited H2O2 production by PMNs after exposure to platelet-activating factor, whereas MHM.23 had only a minimal effect. The effectiveness of different anti-CD18 MAb in preventing reperfusion injury appears to be 1) highly dependent on the specific anti-CD18 MAb employed, 2) predicted only partially by in vitro binding to PMNs, static in vitro tests of PMN adherence, or the extent of inhibition of PMN accumulation in vivo, 3) related more to their ability to inhibit oxidant release from activated PMNs, and 4) strongly influenced by the severity of myocardial ischemia before reperfusion.