saving cost is a significant factor in the successful operation of heavy civil engineering, such as highway and dam construction, and mining projects, which can be achieved through the reduction of operating costs in large projects. in particular, earthmoving operations form a large portion of these projects, and some of the main components of the earthmoving operating cost are fuel, parts, and tires costs, which directly depend on the quality of the unpaved access roads in the field. this study aims at developing a simulation-based model which dynamically estimates the roughness defect score (rds) of the road (performance of the road condition) as the traffic increases and provides an optimal maintenance management program based on the affected cost factors using simphony.net modelling environment. simhony.net is a useful tool for the simulation because it provides an overview of the system’s performance in cyclic and long-term operations. this model uses a stochastic approach to calculate the road resistance and the frequency of maintenance, by considering the variations in nondeterministic variables, such as speed and hauled loads. also, a markov model-based algorithm is being incorporated in the system to provide more realistic modelling of the road deterioration over time. markov modelling involves discrete-event transitions, which model’s road deterioration from a state to another state over time. comparison of these three modelling methods (deterministic, stochastic- monte carlo and stochastic- markov chain modelling) is demonstrated in this study. constant values were used for the deterministic modelling, probability distributions were used for the monte carlo simulation, and transition matrices were used for the markov chain modelling. based on the results, the stochastic modelling was able to provide reliable vehicle operating cost (voc), optimum frequency of maintenance, and road deterioration for ongoing or future cases.