In coronary artery bypass grafting (CABG), arterial grafts demonstrate superior long term patency rates than their venous counterparts. Evidence supports a role for oxidant stress in endothelial dysfunction and it is this dysfunction that precedes the development of atherosclerosis and, more relevant to bypass grafting, intimal hyperplasia. Our aim was to determine whether human arterial endothelial cells (ECs) have a greater ability to protect themselves from oxidant stress than venous graft ECs. ECs were isolated from sections of human radial artery (RA) and human saphenous vein (SV) obtained surplus to CABG, with informed patient consent and ethical committee approval. Following culture as described previously (Conant et al. 2000), cells were treated with 20nM TNFα 24 hours prior to cell extraction and compared with untreated controls. The enzyme activities and total glutathione (GSH) were determined in cell lysates by established methods. Superoxide dismutase (SOD) was determined by the method of McCord & Fridovich (1969). Catalase activity was measured as described by Beers & Sizer (1952) and GSH levels as described by Tietze (1969). Protein levels were determined by MicroBCA (Pierce, USA). Data are presented as mean±s.e.m. The activity of SOD (U/mg protein) was 3.2±0.6 in SVECs (n=6) and 3.9±1.3 in RAECs (n=4). Extracellular SOD, measured as the amount of activity removed by washing cells with excess heparin prior to isolation, was only measurable in RAECs. Cu/ZnSOD or MnSOD activity, measured as the component of activity sensitive or insensitive to 2mM NaCN, respectively, showed no significant difference between the two groups. In SVECs and RAECs TNFα increased total SOD activity 1.6±0.2 times and 1.5±0.1 times, respectively, solely by increases in MnSOD activity. Catalase activity (U/mg protein) was determined as 10.2±1.6 (n=5) in SVECs and 9.7±2.8 (n=4) in RAECs. GSH levels (nmoles/mg protein) were 21.4±4.7 in SVECs (n=5) and 11.1±5.8 (n=4) in RAECs. TNFα did not increase catalase activity or GSH levels in either cell type. In conclusion, extracellular SOD was only found in measurable amounts in RAEC. Neither SVEC nor RAEC showed significant differences in either total SOD or catalase activity or GSH levels. TNFα stimulation increased only MnSOD activity. Extracellular SOD was previously thought to only be produced by vascular smooth muscle (Fukai et al. 2002). The production of extracellular SOD by arterial ECs may confer a first line of defense against extracellular oxidants not available to SVEC.