OBJECTIVES: To estimate the relative efficacy of aclidinium bromide 400 µg BID (AB400), to tiotropium bromide 18 µg QD (TIO18) by means of lung function in patients with COPD, within the first 24 weeks of treatment and illustrate the repeated measures network meta-analysis (NMA) models. METHODS: A systematic literature search using a predefined strategy in MEDLINE, EMBASE and the Cochrane Library identified 16 unique placebo-controlled RCTs reporting FEV1 trough: TIO18 (n=13) and AB400 (n=3). The development of trough FEV1over time for AB400, TIO18 and placebo (PLA) was modeled with fractional polynomials, and the difference between the parameters of these polynomials within a trial were synthesized across studies with a Bayesian NMA. This type of NMA allows for the simultaneous analysis of outcomes at multiple time points. The within-trial correlation was not available from the publications of the included studies, and as such a sensitivity analyses was performed assuming different values for the correlation. RESULTS: Given the fractional polynomial parameters obtained with the NMA model, the corresponding treatment effects over time for AB400 vs TIO18, AB400 vs PLA and TIO vs PLA were estimated. The model with t-0.5and log(t) had the best fit according to the deviance information criterion (DIC). These polynomials and within study correlation were used for the modeling of the outcomes over time. AB400 is equally efficacious compared to TIO18 during the first 24 weeks, as the 95% CrI of the difference in CFB between the treatments includes zero while the mean is <15mL. Furthermore, the probability that each treatment was best was calculated as a function of time. CONCLUSIONS: This analysis demonstrates the use of the proposed NMA models and suggests that maintenance treatment with AB400 results in comparable improvements in lung function, as TIO18 in COPD patients over a 24 weeks period.