Peculiarities of statistical modeling of the transspectral processes used in laser sensing of meteorological parameters of the atmosphere are analyzed. The necessity of using rigorous Monte Carlo technique appears in the case of operation of the corresponding meteorological lidars under cloudy conditions in the atmosphere, when the contribution of multiply scattered radiation to lidar return signal becomes significant. Principal attention in this paper is paid to peculiarities of constructing the algorithms for modeling the processes of the radiation transfer. An important role of the proper choice of the model of elastic and inelastic scattering is demonstrated. In particular, when modeling the angles of photon scattering in the process of the Markov wandering, the scattering phase function shape must be chosen adequate to conditions of the Raman excitation. Results of methodical calculations are presented in comparison with data of other authors. Based on the results obtained, some peculiarities in the formation of multiple scattering background in the Raman sensing channels have been found.