Effects of nonfragmenting solid brass bullets (5.56 mm, 3.1 g, 1.9 cm long, machine-made) and fragmenting soft-point bullets (5.56 mm, 3.2 g, 1.7 cm long, commercially made) were compared when they were fired through soft tissue of the hind legs of 5 live swine (50-70 kg). The swine were anesthetized endotracheally (0.8% halothane) and placed in the supine position with the hind legs extended. Blocks of tissue simulant (10% gelatin at 4.degree. C, molded in blocks 20 .times. 22 .times. 47 cm) were placed against the skin at the predicted point of bullet exit. All shots (a fragmenting bullet through 1 hind leg and a nonfragmenting bullet through the other hind leg of each swine) were fired at a range of 3 m, from a rifle with a bullet tract at 90.degree. to the long axis of the swine''s body. Bullet velocities ranged from 930-990 m/s. Dissections of the bullet tract (through tissue and gelatin) revealed that tissue disruption from the fragmenting bullets was significantly greater (P < 0.001) than from nonfragmenting bullets. The recovered bullets were weighed. Fragmenting bullet lost 59-77% of its original weight and the nonfragmenting bullet was the same weight as originally. Recognition of the amount of tissue disruption and identification of bullet fragments in the wounds resulting from the 2 different bullets should be a useful guide to operating surgeons in selecting the best approach for treatment of [human] gunshot injuries.