The characteristics of gastroesophageal reflux disease have not been adequately defined. To determine the influence on the esophageal mucosa of hiatal hernia, lower esophageal sphincter pressure, acid reflux, and cigarettes and alcohol, we studied the reflux parameters, smoking habits, and alcohol consumption of 184 healthy, ambulatory outpatients who received endoscopy as the initial diagnostic procedure for workup of gastroesophageal reflux. Patients received endoscopic and histologic evaluations of the esophageal mucosa, prolonged ambulatory esophageal pH monitoring, and esophageal manometric determinations. Structural analysis was used to test the plausibility of various clinical theories concerning the most important factors contributing to the development of esophagitis. Statistical analyses revealed the following: (a) the lower esophageal sphincter pressure, acid contact time, and frequency of reflux episodes were highly associated with the presence of a hiatal hernia (p less than 0.003 for all parameters); (b) individuals with esophagitis had 16.5 times as many hiatal hernias as found in normal, healthy people; (c) cigarette smoking was not correlated with esophagitis but was significantly associated with increased lower esophageal sphincter pressure (r = 0.18; p less than 0.03); and (d) smoking was also not associated with increased acid contact time or increased frequency of reflux episodes. We conclude that (a) the presence of a hiatal hernia, not the pressure of the lower esophageal sphincter, is the most important predictor of reflux frequency, acid contact time, and esophagitis; (b) a decreased lower esophageal sphincter pressure, as suggested by structural analysis, is unlikely to be the cause of increased reflux episodes or esophagitis; and (c) if smoking and lower esophageal sphincter pressure are factors in the development of esophagitis, they damage the esophageal mucosa by mechanisms other than increased frequency of reflux episodes or increased acid contact time.