Lack of an endothelial surface is the most important variable causing the relatively poor patency of synthetic bypass grafts. This study was designed to investigate the effect of endothelial cell seeding on small-diameter Dacron grafts seeded with microvascular endothelial cells from omentum, and to evaluate two methods (manual vs automatized) for one-stage seeding in a canine carotid artery model. In 30 mongrel dogs microvascular endothelial cells were harvested from omentum, either by a manual or an automatized method, and seeded onto 6-mm internal diameter Dacron prostheses prior to the graft interposition into the common carotid arteries. Non-seeded Dacron grafts were used as control grafts. All dogs received dipyridamole (75 mg/day) and acetylsalicylic acid (325 mg/day) for 4 weeks. The prostheses were explanted between 2 and 26 weeks after insertion. The results were assessed by patency, angiography, light and scanning electron microscopy, transmission electron microscopy, and morphometry. Endothelial cell seeding improved the patency rate significantly, regardless of the seeding methods used. The overall actuarial patency rates at 5, 12, and 26 weeks were 98%, 94% and 94%, respectively, for the seeded Dacron grafts, and 92%, 62% and 54%, respectively, for the non-seeded grafts. The automatized method yielded more endothelial cells per gram of omental tissue than the manual method (P = 0.0002), but there was no difference (P = 0.34) between the seeding densities per square centimeter of the graft surface. The harvesting and seeding by the automatized method took 55 min for the whole procedure, 20 min less than the manual method. We concluded that one-stage endothelial cell seeding with omental microvascular endothelial cells improved the patency of small-diameter Dacron grafts in a canine model. The automatized method obtained excellent results comparable to the manual procedure, and also reduced the time necessary for the cell seeding.