In a prospective study including 1526 consecutive endoscopies, attempts were made to characterize the benefit of upper gastrointestinal endoscopy. Before endoscopy judgements were made about the most likely diagnosis and treatment and about the degree of suspicion of upper gastrointestinal malignancy. After endoscopy the same types of judgement were made again. The study showed that about half of the endoscopies disclosed clinically significant abnormalities. Furthermore, about every third endoscopy led to unpredicted diagnostic and diagnostic and therapeutic consequences. The benefit was comparably small in patients below the age of 40 years and particularly great in patients above the age of 65, in patients submitted to endoscopy because of barium meal pathology or general suspicion of malignancy, and in patients with upper gastrointestinal bleeding. In general, the present study supports the widespread use of upper gastrointestinal endoscopy in clinical practice.