Small diameter (⩽6mmID) synthetic vascular grafts, used as lower-limb vessel replacements in patients without suitable autologous saphenous veins, have a failure rate of 53% after 4 yr. Graft failure is due to thrombosis and intimal hyperplasia, an increase in smooth muscle cells in the lumen of the vessel which leads to progressive closing and ultimate occlusion of the vessel. In an effort to increase patency rates of synthetic grafts, investigators have seeded vascular grafts with endothelial cells prior to implantation in an attempt to control both thrombosis and smooth muscle proliferation. This technique has been successful for the development of an endothelial monolayer in animal trials, but has met with limited success in humans. The hydrophobicity, low surface energy, and weak electrical charge of expanded polytetrafluoroethylene (ePTFE) provides conditions which are not optimal for endothelial cell attachment. The purpose of this study is to evaluate the effect of vacuum ultraviolet (VUV) modification of ePTFE on endothelial cell adhesion and proliferation. Pieces of ePTFE graft material were exposed to 10, 20 or 40 W VUV radiation for 10, 20 or 40 min using a UV excimer lamp. Prior to cell adhesion and proliferation experiments, the grafts pieces were autoclaved and cut into pledgets. Half of the pledgets were precoated with fibronectin (20μg/ml). Cell adhesion was measured by seeding 3H-thymidine labeled human umbilical vein endothelial cells (HUVEC) onto the pledgets for 60 min. The pledgets were then washed and the remaining radioactivity assayed using scintillation counting. For the cell proliferation experiments, pledgets were seeded with unlabeled HUVEC which were allowed to adhere to the graft material for 18 h. The cells were then exposed to 3H-thymidine (1μCi/ml) for approximately 48 h and then washed to remove any unincorporated 3H-thymidine. Incorporation of 3H-thymidine was measured using scintillation counting. Four replicate samples each, with and without fibronectin, were evaluated for each power and exposure time for both the adhesion and proliferation experiments. VUV modification had no effect on cell adhesion for all power levels studied. In addition, it appears that cell adhesion is independent of the presence of fibronectin. Cell proliferation, on the other hand, is augmented by modification, especially in the presence of fibronectin. These results suggest that VUV modification may provide a better surface for endothelial cell colonization of synthetic vascular grafts.