The containers access and egress to/from the port, on the landside, can be by road or rail modes. Rail transport is used for medium and long-distance trips, while road transport is used for short- or medium-distance trips or if delivery times need to be short. In this last case, it is important to study how drivers of freight vehicles choose their path and how this affects accessibility. This paper presents the most common models proposed in the literature for simulating users’ road path choice behaviour, which belong to the class of random utility models. Starting from path choice estimation, a model that can be used for accessibility evaluation is specified. The paper presents a classification of path choice models belonging to the random utility category, specifying a simple functional form to be applied, i.e. C-Logit. The C-Logit model was first calibrated and validated for freight vehicle movements on the Italian national road network. At the level of path choice, behavioural utility models are used to evaluate satisfaction as a measure of accessibility. In the case of the Logit models, satisfaction can be evaluated in closed mathematical form. A numerical calculation for a small test network is reported.