OBJECTIVES: A network meta-analysis (NMA) evaluating the effectiveness of bimatoprost 0.03%/timolol 0.5% preservative-free (PF) fixed combination (PF BTFC) solution in single-dose vials for the treatment of glaucoma/ocular hypertension (OHT) compared to alternative therapies has been updated (Harvey et al 2013). The outcome of interest was change from baseline (CFB) in intraocular pressure (IOP). METHODS: Systematic searches originally conducted in October 2012 were updated to identify additional randomized controlled trials investigating the efficacy of glaucoma/OHT therapies, where efficacy is defined as IOP CFB. A Bayesian model with random effects for trial, week and time of day was fitted to synthesize the resulting evidence base. RESULTS: In total, 170 studies met the pre-determined mixed treatment comparison inclusion criteria, representing 34 unique treatment arms. PF BTFC was numerically superior to all treatments (monotherapies and combination therapies; preserved and PF therapies) in lowering IOP and statistically significant (p<0.05) for all but seven of the 33 comparisons. Furthermore, despite some trials being adequately sized and the treatment effect showing similar levels of uncertainty to others in the network, the credible intervals (CI) for these treatments were among the widest in the network. CONCLUSIONS: PF BTFC was numerically but not statistically superior to all treatments in lowering IOP. The sample size required to show statistically significant results in NMAs is larger than for head-to-head randomized trials because the precision of comparisons made across a network is proportional not only to the size and uncertainty of individual trials but also to the number of nodes separating treatments (Thorlund 2012). This is thought to be a key driver of the width of CIs for some treatments within this network. Results from the updated NMA will be presented along with a discussion of the interpretation of results for large and complex networks.