The goal of this work is a design and implementation of an on-board driver’s assistance and assessment system. The system overcomes the problem that typical evaluation of skills is performed by experts who may be subjective and are able to consider only a limited number of factors and indicators. The proposed solution is based on eight indicators, which are associated with the vehicle’s speed, acceleration, jerk, engine rotational speed and driving time. These indicators are used to estimate three driving style criteria: safety, economy and comfort. The comprehensive evaluation is done by merging all indicators into one final score. The system is designed according to User-Centred Design method and follows Internet of Things concept. Raspberry Pi minicomputer is used as a central unit to acquire and store the data during the ride and sending them to a server using GSM network. OBD-II interface is used to obtain the data from the vehicle’s network and GPS and accelerometer modules to acquire additional information. MATLAB environment on a local PC is used to process collected data. An outline of the measurements available from ODB-II interface depending on a car model is made. The proposed system has been implemented and evaluated. The evaluation, conducted by collecting readings for specific road actions at different speeds and with different dynamics, confirms that the chosen indicators reliably represent driver’s behaviour. The system was experimentally validated on a group of drivers. The obtained results prove the system’s ability to quantitatively distinguish different driving styles. The system's stability and usability were verified on long-route test. Moreover, the used spider diagram approach established a convenient visualization platform for multidimensional comparison of the result and comprehensive assessment in an intelligible manner. Overall conclusion is that the developed system is a reliable method of the drivers’ behaviour evaluation.