Cholecystectomy is recommended for symptomatic gallstone disease, however atypical abdominal symptoms may be incorrectly attributed to gallstones found by chance at ultrasound. High rates of post-cholecystectomy symptoms confirm that surgery is often performed inappropriately in patients whose complaints do not derive from the biliary tract. We therefore performed esophagogastroduodenoscopy (EGD) in 56 symptomatic patients with gallstones to determine if there was an alternative explanation for patient symptoms. A surgeon then assessed the patient in the light of these findings and decided whether to operate. Of 41 patients who were questioned 12.4 +/- 7.3 months after entry, 1 of 22 patients who underwent surgery, and 5 of 19 patients treated conservatively, were symptomatic (P < 0.05). In the surgical group, one of seven patients with an endoscopic finding remained symptomatic, and in the nonsurgical group one of nine patients remained symptomatic (NS). However, in patients without endoscopic findings, none of the 15 surgical patients but 4 of 10 nonsurgical patients (40.0%) remained symptomatic (P < 0.01). We conclude that patients with gallstone disease and negative EGD who undergo cholecystectomy are less likely to remain symptomatic than those treated conservatively. Thus, endoscopy of the upper gastrointestinal tract in patients with gallstones may provide important information for the surgeon, prior to the making of therapeutic decisions.