BACKGROUND: Optimal long-term management of symptomatic gastro-oesophageal reflux disease (sGERD) patients has not been established. AIM: To determine the clinical value of maintenance intermittent treatment with rabeprazole 20 mg vs. placebo in patients with sGERD. METHODS: This multicentre, US study enrolled patients with sGERD (>or=3-month history of GERD symptoms and >or=4 days/week of heartburn during a 2-week placebo run-in) without oesophageal erosions. Patients with complete heartburn control after 4 weeks of open-label rabeprazole 20 mg daily treatment were randomized to 6-month, double-blind, maintenance intermittent treatment (7- to 14-day courses when heartburn recurred) with rabeprazole 20 mg or placebo. RESULTS: The primary efficacy end point, mean percentage of heartburn-free days, was significantly greater with rabeprazole vs. placebo: 82.58% and 62.17% (ITT; P < 0.0001) [per protocol 86.74% rabeprazole vs. 74.93% placebo (P < 0.0254)]. Compared with placebo group, the rabeprazole group also experienced a significantly higher percentage of heartburn-free daytime (84.06% vs. 63.39%; P < 0.0001) and nighttime (95.41% vs. 90.25%; P = 0.0021) periods, had significantly fewer discontinuations because of insufficient heartburn control (6.3% vs. 36.3%; P < 0.0001) and took fewer antacid tablets daily (0.58 vs. 1.16; P = 0.0021). CONCLUSION: Intermittent use of rabeprazole may be an effective maintenance treatment strategy for patients with sGERD and warrants further investigation. This trial was registered with http://clinicaltrials.gov under the number NCT00165841.