BACKGROUND AND STUDY AIMS: The aim of the present study was to analyze the results of a prospective national survey conducted in Belgium to investigate discrepancies between general practitioners' attitudes and the available guidelines on the management of new patients with symptoms of gastroesophageal reflux disease (GERD). METHODS: A total of 641 general practitioners (GPs) were asked to complete a questionnaire on each consultation by a patient with suspected GERD. RESULTS: The study population included 2234 patients (mean age +/- SD: 47.37 +/- 16.23 years; 52 % aged 45 or older) consulting for the first time for GERD symptoms, without previous endoscopy. Symptoms were classified as mild, moderate, or severe in 39.5 %, 48.0 %, and 11.3 % respectively, and as typical, atypical, or "alarm" in 87.5 %, 16.8 %, and 10.5 % respectively. Upper gastrointestinal endoscopy was requested by GPs in 24.7 % of patients (553 of 2234; including 186 cases that were not justified according to the guidelines), while it should have been requested in 56.6 % (1266 of 2234) if the GPs had respected the national guidelines. Only 28.4 % of the 1171 patients aged 45 or older and 52.6 % of the 95 patients under 45 presenting with alarm symptoms had an endoscopy requested at the first visit. A stepwise multiple logistic regression analysis showed that anemia and symptom severity were the two main factors prompting the GPs to request endoscopy for these patients. CONCLUSIONS: There are major discrepancies between GP practice and national or international guidelines for GERD diagnosis. GPs clearly underused endoscopy in patients with alarm symptoms of GERD and in older patients with a new onset of GERD symptoms. GPs are more strongly influenced by the severity of symptoms than by the type of symptoms.