OBJECTIVES: In the lack of head-to-head comparative trials to demonstrate the efficacy of new treatments, it is common to use network meta-analysis (NMA), including indirect treatment comparison (ITC) or combine direct and indirect evidence through mixed treatment comparison (MTC). Due to the increasing number of drugs approved for the same indication and the increasing complexity of networks for treatments comparisons, new methods of MTC taking into account all the comparisons have aroused. Health Technology Assessment (HTA) bodies increasingly demand NMA although different recommendations about the methodologies to be applied exist. This study aims to review recommendations regarding ITCs and MTCs among the main HTA bodies. METHODS: A review of methodologies for drug comparison recommended by the main HTA bodies was performed. Recommendations related to evidence identification methods, assessment of homogeneity of studies and populations to be combined and statistical approach for the analysis were also reviewed. RESULTS: A systematic literature search is a prerequisite for most HTA bodies, as well as demonstrating homogeneity and consistency among studies. Regarding to the statistical analysis, Bucher’s method is the most commonly used and is recommended by most HTA bodies for indirect comparisons. Nevertheless, some HTA bodies (e.g. HAS, SMC), EUnetHTA and ISPOR Task Force on Indirect Treatment Comparisons consider that even if some direct evidence is available it is appropriate to validate the results using MTCs. According to these institutions, Bucher’s method is not appropriate for the analysis of complex networks, while Bayesian approach is a more comprehensive method that can include meta-regression and study-level covariates. The use of the non-appropriate methods can derive to biased results. CONCLUSIONS: Methodology used for NMA should include all available evidence. Due to the increasing complexity of network patterns, Bayesian analysis better meets HTA needs than the Bucher’s method, and is also a stronger evidence-deriving tool.