Naturalistic driving is an experimentation model that allows us to recognise the driving modes observing the driver's behaviour at the wheel of a set of people in natural conditions during long periods of observation. This research methodology aims at increasing the representativeness of the data collected in opposition to data stemming from highly controlled laboratory experiments. However, naturalistic driving research designs produce large volumes of data that are difficult to handle. Thus, it is very important to work with suitable methods for representing and interpreting data, allowing us to observe the variability of the results. The aim of this study is to implement a new methodology adapted to the particularities of the naturalistic method that allows us to retrieve the positioning information through a georeferencing process of the available data. This method is the first step (preprocessing) to achieve a more clear and intuitive representation (cartographic representation) using Geographic Information Systems (GIS).