Prosthetic vascular replacement materials are still suboptimal especially for distal peripheral and coronary bypass grafts due to the small caliber and high occlusion rates. Therefore new vascular grafts have been assessed with respect to blood compatibility, thrombogenicity and host pretreatment. In 12 heparinized minipigs (25 kg), a total of 120 small caliber grafts (i.d. 3.0 mm) were perfused (6 min) in a femoro-femoral arteriovenous shunt (AVS). Materials included: Dacron (Atrium), PTFE (Gore), Polyurethanes (PU), Auto- and Xenografts. Animals were treated preoperatively with fishoil (3 g EPA/QD x 3 weeks = group I), perioperatively with prostaglandins (PGE1 = group II) or with fishoil and graft incubation in plasma proteins (= group III) and compared to controls without pretreatment (= group 0). We measured flow, deposition of marked thrombocytes (HMPAO-Tc* per area and per weight). - Blood flow was similar in AVS for all groups and materials (168 +/- 86 ml/min, p = NS). There was no significant correlation in deposition of marked thrombocytes and host pretreatments (group 0: controls = 7356 counts/area, group I: fishoil = 6807. group II: PGE1 = 7864). However graft incubation in plasma proteins significantly lowered deposition of marked thrombocytes (1766 in group III vs. 7356 in group 0: p less than 0.05). And significant differences of thrombocyte counts where found between the different materials: Autografts less than PTFE less than Xenografts less than Small Pore PU less than Dacron less than Large Pore PU.--AVS is a useful tool to determine early thrombogenicity in prosthetic vascular replacement materials and to assess the effect of different host pretreatments.