A prospective blinded pilot study was performed at an urban level 1 trauma center to evaluate the efficacy of ultrasound in ruling out penetrating visceral truncal injury. For 8 months, 49 nonconsecutive patients who presented with truncal gunshot and stab wounds were evaluated by a 10-MHz ultrasound tranducer probe. The deepest muscle bundle and the fascia enveloping it was examined by ultrasound. These images were compared to the equivalent contralateral unaffected side of the patient. All the patients then underwent standard testing to evaluate for potential intracavitary injury. Forty-nine patients with a mean age of 28 years (SD, 8.8) were evaluated by ultrasound. A total of 58 injuries were evaluated of which 37 were stab wounds and 21 were gunshot wounds. Thoracoabdominal and back and flank injuries were the most commonly evaluated injuries. There were 20 true positives, 20 false positives, and 18 true negatives, each with approximately twice as many stab as gunshot wounds. There were no false negatives. The sensitivity and negative predictive value of ultrasound in determining clinically significant truncal visceral injury in penetrating truncal trauma is 100 per cent. The specificity and positive predictive value are both approximately 50 per cent. Ultrasonic examination of the injured abdominal wall layers in truncal penetrating trauma is an excellent screening tool. Simple comparative assessment with the unaffected contralateral side allows a highly sensitive method of decreasing the number of potentially morbid, time consuming, and expensive tests that are currently employed to rule out visceral injury.