Pavements are subjected to repeated traffic and environmental actions. These actions lead to the degradation of pavement and affect the pavement performance. The assessment of pavement performance is important not only for selecting pavement design parameters but also for choosing pavement maintenance and rehabilitation strategies. The traffic loads, the environmental actions, and the properties of pavement material are uncertain. These uncertainties must be taken into account when predicting the pavement performance. Since the traffic and environmental actions vary with time, they should be modeled as stochastic processes. In this study, stochastic models of the net traffic load growth and environmental actions are proposed based on the rectangular pulse processes. These models can be used in conjunction with the Ontario Pavement Analysis of Cost (OPAC) model to predict flexible pavement performance in a probabilistic framework. The prediction of pavement performance for a typical pavement was carried out by using the simple simulation technique and the first-order reliability method. The analysis results suggest that the predicted pavement serviceability measured in terms of the riding comfort index depends on the correlation between traffic effects in each year and the correlation between the environmental actions in each year. The results also suggest that the uncertainty in the riding comfort index is controlled by the uncertainty in traffic and environmental actions in the early stage of service, while it is dominated by the material property of the subgrade in the later stage of service.Key words: deterioration, reliability, pavement, serviceability, stochastic process.