Although antireflux surgery is effective for selected patients with gastro-oesophageal reflux disease (GERD), some patients require further antireflux medication after a variable period. We have previously reported that 14% (95%CI 9% to 20%) of patients require antisecretory medication after antireflux surgery (ARS), based on a meta-analysis of 18 RCTs.1 Since most surgical studies examining the efficacy of ARS are not randomised, we aimed to investigate the need for further ARM after ARS in recent non-randomised studies. MEDLINE and EMBASE databases were searched from 2005 up to Oct 2009 for published, English-language, non-randomised studies investigating the efficacy of ARS for GERD in adults, and reporting the use of ARM after ARS. The proportions of patients using ARM after ARS were sample size weighted and pooled by applying a random-effects model using StatsDirect 2.7.2®. In total, 24 clinical studies met inclusion criteria with follow-up ranging from 0 to 3 months to >10 years and most were single arm observational studies. Some studies reported data for more than one time point. Most surgical interventions were fundoplication, while a few studies only stated “anti-reflux surgery”. When all studies were pooled, regardless of the follow-up time point and only considering the last measured data for the same study, 20% (95% CI 13% to 28%) of patients required ARM after ARS (25 arms, n=7739). The corresponding data for the use of PPIs were 18% (11–25%, 20 arms, n=5994). When subgroup analyses were performed for different follow-up periods, the proportions of patients who required ARM after ARS were 15% (8–23%, 5 arms, n=1680), 24% (12–37%, 7 arms, n=1301), 18% (9–30%, 12 arms, n=5929), and 27% (18–37%, 5 arms, n=1447) for <1 year, 1–3 years, 3–5 years, and 5–10 years, respectively. The corresponding data for the use of PPIs were 14% (6–24%, 4 arms, n=836), 36% (18%–56%, 3 arms, n=225), 12% (3–25%, 9 arms, n=3572), and 21% (15–28%, 5 arms, n=1447). After antireflux surgery, 20% of patients still require antireflux medication in non-randomised studies, and this proportion increases with the duration of follow-up to 27% after 10 years. This figure from non-randomised studies, which is higher than that from randomised studies, is probably more representative of the unmet need in clinical practice. However, some important issues remain unanswered, including the reasons for taking ARM after ARS, and the effect of loss to follow-up in long-term observational studies. Furthermore, how best to manage reflux patients who fail surgical treatment requires further investigation.