Three groups each of 16 pigs were fed for 70 days on a diet containing one of three test fats/oils (anhydrous milkfat, fish oil (MaxEPA T M ) or hydrogenated coconut oil) at a level of 100 g/kg. On day 70, a balloon catheter was inserted under anaesthesia into each femoral artery and inflated to 620.5 kPa for five 30 sec intervals. One h after angioplasty, while still anaesthetised, the pigs were killed by exsanguination. The injured segments of femoral artery were perfused in vivo with 0.9% saline and 10% neutral buffered formalin before being harvested. Thrombus size was determined by the counts of technetium-99m-labelled platelets in the injured arteries and by morphometry of the thrombus area in microscopic sections. Of 91 arteries subjected to angioplasty, mural thrombi were seen macroscopically in all nine arteries that had deep injury with exposure of the tunica media, as well as in 19 of the remaining 82 arteries (23%) in which there was no evidence of a medial tear. Many small platelet-rich thrombi, attached to the denuded endothelium, were also detected microscopically. There was no statistically significant effect of diet on either thrombus area or platelet deposition at sites of arterial injury. Failure to show significant differences may have been due to the large within-group variation in thrombus size, a factor which may limit the use of this model in studies aimed at detecting differences in thrombogenicity between different dietary fats and oils.