Four hundred and thirty patients with grade 2 or 3 esophagitis were treated after 2/1 randomization for 8 weeks with omeprazole 20 mg (n = 294) or ranitidine 150 mg bid (n = 136). Apart from treatment, 8 epidemiological factors (gender, age, occupation, obesity, smoking, alcohol, NSAID, and coffee or tea consumption), 5 clinical factors (day/night pain distribution, burning score, severity of regurgitation and of dysphagia, number of painful episodes requiring prescription of an antisecretory agent during the previous year, and onset of symptoms before age 30) and 3 endoscopic factors (grade and upward extension of esophagitis, and existence of hiatal hernia > or = 5 cm) were analysed. The influence of these factors on healing at 8 weeks and on changes in symptoms was evaluated by multivariate analysis. 92.1% of patients enrolled were analyzed. In comparison with ranitidine, omeprazole increased the percentage of healed patients (93% v. 67.5%, p < 0.001) and the rapidity of disappearance of symptoms (5 days v. 7 days, p < 0.001). Independent good prognostic factors associated with healing rate were treatment with omeprazole (p < 0.001) and grade 2 esophagitis (p < 0.001) while those associated with the disappearance of symptoms were a low burning score (p = 0.001), advanced age (p = 0.004), treatment with omeprazole (p = 0.005), the absence of any occupation (p = 0.01) and male gender (p = 0.017). The results of this study show that, apart from treatment, endoscopic factors are predictive of the healing of reflux esophagitis treated by antisecretory agents while clinical factors are more important with regard to the disappearance of symptoms.