This paper analyses convergence in per capita gross regional product of Russia’s regions during the period 1995-2010, when regional data are available. Using a panel regression framework we find no evidence for beta-convergence. Instead we find divergence, which is, however, attenuated over time. Robustness checks that use regional real income instead of gross regional product confirm this outcome as do non-parametric estimates of convergence, namely estimates using Markov transition probability matrices and stochastic kernel plots of regional relative income. Decompositions of regional income and gross regional product also find no sigma-convergence of Russian regions. These decompositions point to the geographical concentration of extractive activities in the Urals and of business services and of the public administration in the Moscow area as the main culprit for this lack of convergence. They also establish that despite reforms to equalize provisions of public goods across Russia, the social services sector of the public administration, education and health still do not have the expected equalizing impact on regional income.