Forecast of transport mobility of population is an important landmark in the development of passenger transport of JSC "Russian Railways". Business processes of transport planning, justification the traffic volume and train composition schemes, the definition of the operating fleet of cars and infrastructure needs are based on the data on projected traffic volume and the dynamics of their changes. The article describes the results of research work on modeling and forecasting passenger flow and the choice with reference calculation an optimal model having the smallest prediction error. Taking into account that the public demand for transport services depends on many conditions, a list of factors was justified and the most affecting ones were selected. The stages of the modeling process by extrapolating trends and multivariate analysis are presented, reflecting the connection of passenger flows with the economic, demographic and regional characteristics. The accuracy of prediction was made together with conclusions about the performance of the proposed models. The correctness of the developed mathematical apparatus for the forecast of passenger flows is confirmed by supervisory calculations. They are based on actual data on passenger railway transport and assessment of share participation of aviation and road transport in the development of demand.