Twenty-five adult patients with liver cirrhosis, and another 30 patients with no liver disease but referred with symptoms suggestive of gastroesophageal reflux disease were selected at random. Twenty-four hour ambulatory intra-esophageal pH measurement and upper gastrointestinal endoscopy were carried out on all patients recruited. Applying the former test, 16 (64%) of the patients with liver cirrhosis have gastroesophageal reflux disease. This figure is comparable with the 70% (21/30) rate recorded in the group of dyspeptic patients clinically thought to have the disorder. A positive endoscopic diagnosis was much lower at 12% and 23%, respectively. No significant differences were observed among liver disease patients when they were subdivided in accordance with the etiology of liver cirrhosis and the grade of esophageal varices. We conclude that gastroesophageal reflux disease occurs at a high frequency (64%) in patients with liver cirrhosis and portal hypertension, irrespective of the etiology of cirrhosis and the grade of esophageal varices. It is therefore considered to be the main cause of esophagitis in these patients, and that it might play a role in initiating a variceal bleeding episode. The latter hypothesis needs further evaluation.