OBJECTIVES: Data supporting the use of the proton pump inhibitor (PPI) test as a diagnostic test of gastroesophageal reflux disease (GERD) in Asia are lacking. The aims of this study were to evaluate the diagnostic characteristics of the PPI test and factors associated with responsiveness to PPIs. METHODS: The PPI test was evaluated using lansoprazole (30 mg bid) for two weeks. All patients underwent endoscopy after symptom assessment. Patients without erosive esophagitis (ERD) underwent 48 h esophageal Bravo pH monitoring. Subjects were considered to be responsive to PPIs if they reported a > 50% reduction in reflux symptom score. RESULTS: Seventy-three patients (M:F = 40:37, 47 +/- 13 years) were enrolled. We identified 46 patients with ERD, 18 patients with nonerosive reflux disease (NERD) and 9 patients without GERD. The PPI response rate was higher in patients with GERD than in patients without GERD (49/64, 77% vs. 4/9, 44%; p < 0.05). The sensitivity, specificity, and positive and negative predictive values were 77%, 56%, 92% and 25%, respectively. The PPI responsiveness was 80% (37/46) in the ERD group and 67% (12/18) in the NERD group. PPI response was not affected by age, sex, Helicobacter pylori, the psychological characteristics or cytochrome P2C genotypes. ERD and symptom-reflux association were the factors affecting PPI responsiveness. CONCLUSIONS: The PPI test was modestly sensitive and specific for diagnosing GERD. However, it would be useful for discriminating patients with ERD. In the NERD group, patients with positive symptom-reflux association would be most benefit from PPI treatment.