SIR-model is one of the most common in the modeling of infectious diseases. This model describes with differential equations the change in the number of three groups of individuals: susceptible, infected and removed from the group of infected. This article focuses on the use of the SIR model in predicting the incidence of HIV infection in Russia. The search for the coefficients of the model is considered in detail: an analysis of the statistical data on HIV infection was made, approximate values of the coefficients were obtained. Prospects for different errors in statistical data are considered.