This chapter illustrates some of the great potential of the stochastic process approach for studying complex policies applied to transportation networks, and considers some of the general issues that arise in the interpretation of the outputs produced. Traffic assignment methods could fulfil at least two major needs in an information setting: short term forecasting of the evolution of congestion patterns, providing predictive information for broadcasting by travel or route information systems; and long term prediction of the operational and economic benefits of information provision. The traditional equilibrium approaches, whether departure time dependent or independent, and whether 'stochastic' or deterministic, are fundamentally based on the premise that no day to day variability exists in the traffic system. One of the major benefits of a stochastic process approach is that it recognizes traffic conditions to be variable, with the effect of this uncertainty on driver behaviour being endogenous to the model.