Conventional cholescintigraphy (60 patients) and a modified protocol (59 patients) were compared in 74 females and 45 males with acute cholecystitis. In the modified protocol, intravenous morphine (0.04 mg/kg) was administered whenever the gallbladder was not seen 40 minutes after injection of Tc-99m-pyridoxylideneglutamate (36/59). Accuracy was 98% with morphine, compared with 88% for the conventional protocol; specificity improved from 83% to 100% with no loss of sensitivity (96% in both groups). Low doses of morphine are well tolerated and can result in a highly accurate diagnosis of acute cholecystitis without the need for delayed imaging.