2 measurement sets were performed by the INRETS (Institut National de Recherche sur les Transports et leur Sécurité - National Research Institute on Transport and their Safety) on vehicles under actual use conditions: 35 gas-engined passenger cars, selected amongst 7 French representative models, manufactured between 1979 and 1981, from 6 French geographical areas, were studied from 1983 to 1985, 20 gas-engined vehicles selected amongst two recent models widely sold between 1988 to 1989, in the Paris area and in the provinces were studied between 1989 and 1990. In both cases, the vehicles selected were private cars, equipped with on-board measurement and data acquisition systems and sensors. They were driven by their owners in the course of their day by day use. Vehicle and engine speeds, throttle position, fuel consumption, engine oil and water temperatures, as well as uses or operation of the manual choke, headlamps, wipers, rear-window heater and the engine cooling fan were recorded at second time intervals. During the second experimental set of measurements, outside ambient temperature, power consumption of the vehicle electrical system and brakes use were recorded too.10,000 trips were recorded for 71,000 kilometers covered which represented 1,260 hours of vehicle running. This significant amount of data allowed obtaining a very accurate statistical basis as relates to the actual use of vehicles.A short description of the method used is given. Vehicle uses and running conditions are analyzed through all the data recorded, taking into account the geographical areas, the vehicle models, the drivers and the trip distribution according to time: The vehicle uses: frequency of daily uses, trip durations and distances covered, trip distribution within the day or the week, The trip characteristics: duration and length, speeds used, number of intermediate stops, etc… The engine and vehicle running conditions: engine speeds used, engine load, gear changes, running thermal conditions, fuel consumption, etc… A great stability can be observed between the two sets of measurements (average trip of 11 min. 30 for 7 kilometers covered and of 10 min.30 for 7.3 kilometers covered, for the first and second set respectively; with very similar duration and distance distributions), but a significant increase of vehicle daily uses can be noted. Such a stability confirms the validity and representativity of these measurements, even if they were carried out on a limited sample. Use distributions show a significant number of very short trips.