OBJECTIVES: Guidelines for treatment of rheumatoid arthritis (RA) advise initial therapy with non-biological disease modifying antirheumatic drugs (DMARDs). In case patients do not respond adequately, treatment should be switched to biologic DMARDs.  Aim of this research is to compare results of public available systematic reviews (SRs) on comparative effectiveness (CER) and potential impact of differences in methodology. METHODS: We performed literature research for SR on CER of biologics for the treatment of RA. Search was limited to reviews published in 2009 or later. Methods of the reviews and results were extracted from the publications. Results are summarized in narrative way and differences in results are reflected focusing on methodological key issues. RESULTS: Eleven recent SRs were identified addressing the question of CER of biologics. Since there are no head-to-head comparisons available for all but one biologic, reviews had to use indirect comparisons to assess CER. Authors used the Simon-Bucher approach or Bayesian methods (mixed treatment comparisons). Clinical trial guidelines for RA give the advice to use Response Criteria defined according to the American College of Rheumatology (ACR) as primary parameter. Therefore, all trials assessed ACR20 (20% improvement), ACR50 and ACR70 and could be used for comparison. Further parameters were not assessed in a uniform manner (e.g. quality of life) or not assessed in all trials (e.g. disease activity score) and therefore could not be used for indirect comparison of treatments. CONCLUSIONS: Due to lack of head-to-head data for comparison of biologics, statistical methods for indirect comparison have to be used to answer the question of CER. These methods have restrictions and base on assumptions that might be heavily violated.  Substances were tested over a time period of more than 10 years with effects on study population and variation in study designs. Nevertheless, the results seem to be fairly consistent.