BACKGROUND Focused Assessment with Sonography for Trauma (FAST) is rapidly establishing its place in the evaluation of blunt abdominal trauma. However, no prospective study specifically evaluates its role in penetrating abdominal trauma. METHODS Data were collected prospectively in 75 consecutive stable patients with penetrating trauma to the abdomen, flank, or back, from December 1998 to June 1999. Those with an obvious need for emergent laparotomy were excluded. FAST was performed as the initial diagnostic study on all patients. Wound location, type of weapon, and findings of diagnostic peritoneal lavage, triple-contrast computed tomographic scan, or laparotomy were recorded. The presence of peritoneal blood was noted. Data were analyzed using the chi(2) test. RESULTS Of the 75 patients, there were 32 stab and 43 gunshot wounds. There were 66 male patients and 9 female patients; the mean age was 30 years; 41 had proven abdominal injury and 34 had no injury; and 21 patients had a positive FAST. Nineteen had peritoneal blood and injuries requiring repair at the time of laparotomy. There were two false-positive studies. Fifty-four patients had a negative FAST. In 32 patients, this was a true-negative study. Thirteen patients had a false-negative FAST and had peritoneal blood and significant injury on further evaluation. Nine patients had a negative FAST and no peritoneal blood but still had abdominal injuries requiring operative repair, including liver (four), small bowel (four), diaphragm (three), colon (three), and stomach (one). The overall sensitivity of FAST was 46% and the specificity was 94%. The positive predictive value was 90%, and the negative predictive value was 60%. CONCLUSIONS FAST can be a useful initial diagnostic study after penetrating abdominal trauma. A positive FAST is a strong predictor of injury, and patients should proceed directly to laparotomy. If negative, additional diagnostic studies should be performed to rule out occult injury.