Although the effect of short-term treatment on gastroesophageal reflux disease (GERD) has been studied, there is little information concerning the long-term outcome of this disease. The aim of the present study was to follow up a patient population with GERD to identify variables that could be used to predict the disease course. From 1989 to 1991, we assessed 349 patients with GERD. Barium roentgenogram, esophagoscopy with biopsy, and ambulatory esophageal 24-hour pH monitoring were performed in each patient at baseline. One hundred eight outpatients with typical symptoms of gastroesophageal reflux and pathologic findings on 24-hour pH monitoring were included in the study. After diagnosis of GERD, patients were instructed to adhere to a specific diet and lifestyle regimen. Subsequently, on the basis of symptoms, the severity of esophagitis, and the degree of acid reflux, patients were assigned to a pharmacologic treatment using antacids, cisapride, ranitidine, and omeprazole, alone or in combination. Patients were followed up after 40 days and then yearly for 7 years using endoscopy and clinical examination. During the 7-year follow-up, 14 patients (13%) dropped out (2 of the 14 died of causes not related to GERD) and 6 patients (6%) underwent surgery to correct the reflux. Twenty-three patients (21%) discontinued pharmacologic treatment because of a stable improvement of their symptoms, whereas 65 (60%) continued the drug therapy. The present study demonstrated that reflux symptoms can improve and even disappear after treatment, but only in a small percentage of patients with GERD. The majority of patients must continue drug therapy. Moreover, a complete evaluation of patients at entry showed the following factors to be highly predictive of unfavorable disease progression: the presence and grade of esophagitis, a high grade of gastroesophageal reflux during 24-hour pH monitoring and while in the supine position, and the presence of hiatal hernia