OBJECTIVES: The questionnaire of the Cochrane Collaboration developed to assess risks of six key systematic biases in randomized controlled trials (RCT) is needed for wide use in Russia, so it was translated into Russian and validated. METHODS: Two experts (F.V.K. and K.G.R.) with 4-year experience of evidence assessment used the Russian version of the questionnaire to consider biases in 20 RCT published in 2002-2012. The inter-rater agreement was estimated using Kappa (K) coefficient, it was calculated for six key domains and also for overall summarizing risk of bias (low, high, unclear) within a study (across domains).  RESULTS: Two experts demonstrated good, very good and excellent agreement on each key domain and also good overall summarizing agreement (K=0.767, 95% CI 0.527; 1.000). We propose to consider the conflict of interests not as additional but as obligatory (the 7^th) key domain (the agreement in this domain is very good, K=0.840, 95% CI 0.641; 1.000). As it appeared to influence greatly the overall estimation summarizing risk we also propose to modify slightly the Cochrane rule of summary assessment of the risk of bias (across domains). The modified rule allows unclear risk of bias in one key domain to have low overall summarizing risk. The final version of the document also provides good agreement (K=0.785, 95% CI 0.503; 1.000). CONCLUSIONS: The Russian version of the Cochrane questionnaire is ready to use also as a web-based tool.