PURPOSE: To determine the most helpful abdominal CT findings in patients with small bowel perforation withoutpneumoperitoneum or oral contrast leakage after blunt trauma. MATERIALS AND METHODS: We retrospectively analyzedthe abdominal CT findings of 51 patients with small bowel perforation without pneumoperitoneum or oral contrastleakage. A score of 2 was assigned if bowel wall thickening of more than 5 mm or enhancement of the bowel wall wasdefinite, and 1 if equivocal thickening or enhancement of the bowel wall, mesenteric infiltration, sentinelclotting, intermesenteric fluid, or ileus were observed. According to the score, each finding was classified asstrongly positive(5-8), possibly positive(3-4), or probably negative(1-2), and the accuracy of each classificationwas evalvated. We decided which findings would be most helpful when attempting to diagnose small bowelperforation. RESULTS: Forty one bowel perforations and four mesenteric injuries were identified in 45laparatomies among 51 patients. In all 20 patients with a score of more than 5, small bowel perforation was foundon surgery. The prevalence rate of bowel wall thickening and enhancement was significantly different betweenpatients with perforation(88%, 80% respectively) and those without (20%, 40% respectively)(p<0.05). Thesensitivity, specificity and accuracy ; of each criterion were compared (sum of scoring of all CTfindings > or = 3:82.9% 40% 74.5%, scoring of bowel wall thickening and enhancement > or = 2: 82.9% 80% 82.4%), and itwas found that for specificity and accuracy, the latter was superior to the former. ROC analysis using the scoresof each criterion as cut-off value also showed that the curve of the latter showed a more marked upward trend andwas the most helpful parameter. CONCLUSION: When diagnosing small bowel perforation in the absence ofpneumoperitoneum or oral contrast leakage after blunt abdominal trauma, bowel wall thickening and enhancement weremore helpful parameters than the sum of all CT findings.