This paper investigates classifications of road type and driving style based on on-board diagnostic data, which is commonly accessible in modern vehicles. The outcomes of these classifications can be utilized in, for example, supporting the advanced driver assistance systems (ADAS) for enhancing safety and drivability, and online adaptation of engine controller for improving performance and fuel consumption. Furthermore, the classifications offer valuable information for fleet operators to consider when making decision on procurement plans, maintenance schedules and assisting fleet drivers in choosing suitable vehicles. To this end, a velocity-based road type classification method is evaluated on measurements collected from real driving conditions and compared to an open-sourced map. To produce representative results, two most commonly adopted driving style classification methods, i.e. acceleration and jerk-based methods are evaluated and compared on the same set of measurements. The classification results and their correlations with fuel consumption are also investigated and discussed. This investigation reveals that the acceleration and jerk-based driving style classifications are only applicable to certain driving conditions, prompting for the need of a more comprehensive classification of driving style.