We assessed bowel sounds in controls and patients with acute abdominal conditions using a new computer-aided sound analysis system (C.A.S.A.S.). Sixty-three controls and 61 patients with an acute abdomen had 10-min recordings of bowel sounds in a silent environment with computer analysis of the digitized acoustic signal. Mean bowel sound length was 20 +/- 1.3 ms in controls, 32 +/- 30 ms in cholecystitis, and 37 +/- 35 ms in intestinal obstruction (p < 0.05; mean +/- sd, Mann-Whitney U test). Similar significant differences existed between controls and acute abdomens in the number of sounds generated, sound intensity, and intervals between sounds. Although they were significantly longer in intestinal obstruction, bowel sounds were reduced in number (0.6 sounds/s) as compared to controls (0.4 sounds/s; p < 0.05). Significant differences were seen in all acoustic parameters in patients stratified for the presence or absence of peritonitis. C.A.S.A.S. allowed an objective scientific assessment of bowel sounds and identified significant differences between bowel sound patterns in controls and patients with an acute abdomen.