In this paper, we propose a finite-state Markov model for per-user service of an opportunistic scheduling scheme over Rayleigh fading channels, where a single base station serves an arbitrary number of users. By approximating the power gain of Rayleigh fading channels as finite-state Markov processes, we develop an algorithm to obtain dynamic stochastic model of the transmission service, received by an individual user for a saturated scenario, where user data queues are highly loaded. The proposed analytical model is a finite-state Markov process. We provide a comprehensive comparison between the predicted results by the proposed analytical model and the simulation results, which demonstrate a high degree of match between the two sets.