Summary Background Gastro-oesophageal reflux disease (GERD) is characterised by symptomatic heartburn and regurgitation. Treatment with proton pump inhibitors (PPI) effectively decreases heartburn symptoms, but their effects on symptomatic regurgitation are less clear. Aim To determine the impact of PPI therapy on heartburn and regurgitation severity in patients with either non-erosive GERD (NERD) or erosive oesophagitis (EE). Methods Endoscopically-confirmed NERD patients received dexlansoprazole 30 or 60 mg or placebo in a randomised, blinded, 4-week study. Endoscopically-confirmed EE patients received dexlansoprazole 60 mg or lansoprazole 30 mg in two 8-week, randomised, blinded healing studies. The Patient Assessment of Upper Gastrointestinal Symptom Severity questionnaire, which includes a heartburn/regurgitation subscale, was administered to assess symptom severity at baseline, and at weeks 2 and 4 of the NERD study and at weeks 4 and 8 during the EE trials. We defined separate subscales for heartburn and regurgitation for this post-hoc analysis. Among patients with both symptoms at baseline, improvements in individual heartburn and regurgitation subscales along with the original combined heartburn/regurgitation subscale were determined. Results In the NERD and EE studies, 661 and 1909 patients, respectively, had both heartburn and regurgitation at baseline. NERD patients receiving dexlansoprazole 30 and 60 mg experienced significantly greater improvements in symptom severity for both heartburn and regurgitation compared with placebo. EE patients receiving dexlansoprazole 60 mg had significantly greater improvements in heartburn/regurgitation and heartburn-only subscales at week 4 compared with those receiving lansoprazole. Conclusions Dexlansoprazole appears to be effective in improving both heartburn and regurgitation, and this improvement is maintained for the duration of treatment.