Objectives To evaluate the accuracy of emergency physicians using bedside ultrasound to detect appendicitis (BUSA). Methods Patients presenting to the emergency department with a clinical suspicion of appendicitis were prospectively enrolled and received a 5-min BUSA. Patients received routine work-up for acute appendicitis as deemed appropriate by the attending physician. Radiologists and consulting surgeons were blinded to BUSA results. The criterion standard for the presence or absence of acute appendicitis was the pathology report for patients who received appendectomies, and telephone follow-up for patients discharged home without surgical intervention. Results A total of 132 patients were enrolled. In 44 cases BUSA was positive. Of these, 37 had surgical pathology reports consistent with acute appendicitis, whereas seven did not have appendicitis. In 82 cases, BUSA was negative. Of these, 62 were determined not to have appendicitis, whereas 20 had appendicitis by pathology. Sensitivity for BUSA was 65% [95% confidence interval (CI) 52-76], specificity was 90% (95% CI 81-95), positive predictive value was 84% (95% CI 71-92), and negative predictive value was 76% (95% CI 65-84). The likelihood ratio of a positive BUSA was 6.4 (95% CI 3.1-13.2). Five patients discharged home with a diagnosis other than appendicitis were unable to be reached by telephone, and were excluded from data analysis. Conclusion Our study gives insufficient evidence to support the use of bedside ultrasound by emergency physicians to rule out appendicitis. The high specificity in our study, however, suggests that with further training, BUSA may be useful to rule-in appendicitis in some patients. © 2008 Wolters Kluwer Health Lippincott Williams & Wilkins.