Résumé : AIM: We conducted a prospective survey of the use, relationship to existing guidelines and outcome of empirical treatment by general practitioners (GPs) in gastroesophageal reflux disease (GERD) patients. METHODS: In total 815 GPs completed a questionnaire on consecutive GERD patients in whom they started empirical therapy without additional investigation. RESULTS: The study population included 4,177 patients (50.3% men, mean age 39.3 +/- 11.0 years) without previous endoscopy presenting with typical GERD symptoms (96%). According to Belgian-French guidelines, endoscopic examination was recommended in 2,083 patients, but all received empirical therapy. Data were collected for 3,955 patients (94.7%) at 1-month follow-up; 3,360 patients (85.0%) were responders to empirical therapy (no symptoms during the last 7 days), which was not significantly influenced by the type of treatment. Presence of alarm symptoms (15.5%), alcohol consumption, failure of previous treatment or recurrence after previous treatment were significant predictors of nonresponsiveness to empirical therapy. Follow-up endoscopy in 521 patients showed reflux esophagitis in 70.6%, peptic ulcer in 5.9% and Barrett's in 0.2%. CONCLUSIONS: Empirical treatment of typical symptomatic GERD by GPs is highly successful. Predictors of nonresponsiveness to empirical therapy are the presence of alarm symptoms, alcohol consumption and failure of or recurrence after a previous treatment.