Road traffic and its influence over individuals is an important aspect of our life nowadays. Its study in order to understand its dynamics and the factors that affect it is a relevant field of research. Traffic simulations have become a fundamental tool for these studies. They provide a controlled environment to analyse traffic settings. However, they present some shortcomings. One of the main ones is the need of multidisciplinary groups of experts to work with complex models. Communication problems and misunderstandings frequently appear in them, which produce mistakes and bring increased costs. Some works have addressed these issues adopting abstract concepts that can act as bridges among different groups to model and implement simulations. Works that use intelligent agents to represent individuals, and their related simulation platforms, belong to this category. Nevertheless, these platforms are still programmer-oriented, so other experts find difficult to ground their abstract models in them. As a further step, Model-Driven Engineering (MDE) has been proposed to work with models and simulations. It offers the possibility of working with models at multiple levels of abstraction and focused on different aspects. These models can be oriented to specific experts’ backgrounds. The work presented follows this approach and introduces a generic Modelling Language (ML) through a model, that can be specialized to meet different needs in road traffic simulations. The case study illustrates how that model can be successively modified to model people’ behaviour in traffic both at the traffic expert and platform-oriented levels. This allows reducing the learning curve of experts with backgrounds non-related to software simulations.