This paper considers different ways of computing indexes for forecasting economic activity in Russia. The first is the methodology used by the Russian Development Centre based on the concept of "growth cycles". The second combines the dynamic principal components and dynamic factor analyses. The third approach is the NBER methodology based of diffusion indexes constructed using a dynamic factor model. This paper is an attempt to reveal strengths and weaknesses of the three methods in application to Russian data and to develop a better methodology for forecasting economic activity in Russia.