The conclusions drawn from a prospective survey of 500 patients examined by cholecystography (OCG) and ultrasound (UCG) are presented. First, both procedures were found to be highly accurate in detecting calculi, with false-negative rates of 1%. Secondly, if OCG is abandoned in favour of UCG, most acalculous adenomyomatosis (and many polyps) will not be diagnosed. If acalculous as well as calculous disease is regarded as clinically important, fluoroscopic OCG is the examination of choice. However, if a clinician wishes solely to find or exclude stones the investigations are equally accurate. Thirdly, since some calculi and some acalculous disease will be missed by either technique, serious consideration should be given to the further investigation of all gallbladders considered normal by either examination. A protocol is described which, though unorthodox, enables a decision on the status of the gallbladder to be made on a single visit to the radiology department.