Objective: Numerous randomized controlled trials (RCTs) have evaluated efficacy of proton-pump inhibitors (PPIs) in controlling nighttime symptoms of gastroesophageal reflux disease (GERD). Quantitative synthesis of the effect of PPIs on nighttime symptoms is lacking, thus the validity of performing a meta-analysis was assessed. Methods: MEDLINE and EMBASE databases from 1990-June 2007 were systematically searched for RCTs evaluating the efficacy of PPIs on nighttime symptoms in adults with GERD. Methodological and clinical homogeneity across studies was explored. Methodological diversity or differences in study quality evaluated by Jadad score (ranging from 1=low to 5=high) and clinical diversity in nighttime criteria used for patient enrollment, nighttime outcomes measured, and the nighttime definition used were explored. Results: Thirty-two RCTs compared the efficacy of PPI with placebo only (n=7), H2-receptor antagonist only (n=12), another PPI only (n=11) or both placebo and H2-receptor antagonist (n=2) in controlling nighttime GERD. The majority of studies (n=28) were of high methodological quality (Jadad score of at least 3 points). Source of data collection was patient daily diaries across all studies. Criteria for enrolling nighttime GERD patients (frequency and/or severity of nighttime symptoms) lacked consistency. Nighttime heartburn measures varied from percentage of patients without heartburn (n=18), percentage of heartburn-free nights (n=15), heartburn severity score (n=11) or time to heartburn relief (n=6). Most studies assessed efficacy at eight weeks or less; only three studies measured the long-term efficacy of PPI. However, very few nighttime heartburn measures assessed the same timeframe. The time window for the nighttime symptom assessment was reported in only three studies and was not based on specific hours but on sleep/posture (retiring/lying down to sleep). Conclusion: RCTs of PPI therapy in nighttime GERD are of high methodological quality. However, presence of clinical heterogeneity across trials in enrollment criteria, outcomes, and timeframes minimizes the possibility of performing meta-analysis.