This paper presents initial results from a series of eye-tracking experiments on a Full Mission Bridge simulator. The aim of this research was to develop a stochastic model of a navigator’s attention distribution during their navigational watch. Such model could be used as a tool for workload and usability studies for navigators and navigational equipment interfaces. A structure of the model is discussed together with the evaluation of Markov chains as a main modelling tool. Initial results are presented and discussed. It is suggested that 1st order Markov chains are not fully applicable for this problem. A combination of the 1st and higher-order Markov chains will be applied in the next stage of research.