The diagnosis of appendicitis is frequently difficult. We studied prospectively the diagnostic accuracy and clinical impact of abdominal ultrasonography in 111 consecutive patients thought to have appendicitis. Ultrasonography was performed with small high-resolution, linear-array transducers, with the abdomen compressed to displace or compress bowel and fat. Among 52 patients later shown in surgery to have appendicitis, ultrasonography was unequivocally positive in 39 (sensitivity, 75 percent). Of 31 patients in whom appendicitis was definitely excluded, none had a positive ultrasound examination (specificity, 100 percent). The sensitivity in those with a perforated appendix (28.5 percent) was much lower than in those with acute nonperforating appendicitis (80.5 percent) or appendiceal mass (89 percent), but the low sensitivity did not influence clinical management, since the need for surgery in patients with a perforated appendix was clinically obvious. Ultrasonography resulted in changes in the proposed management in 29 of the 111 patients (26 percent). It also led to the correct diagnosis in the 16 patients who were found to have a disease other than appendicitis. We conclude that ultrasonography is a useful aid in the diagnosis of appendicitis.