In-use fuel consumption data of 924 passenger cars (611 petrol, 313 diesel) were collected from various European sources and were evaluated in comparison to their corresponding type-approval values. The analysis indicated that the average in-use fuel consumption was higher than the type-approval one by 11% for petrol cars and 16% for diesel cars. Comparison of this dataset with the Travelcard database in the Netherlands showed that the deviation increased for late model years and in particular for cars with low type-approval values. The deviation was higher than 60% for vehicles registered in 2012 within the 90–100 gCO2/km bin. Unrealistic vehicle resistances used in type-approval were identified as one of the prime reasons of the difference. A simplified linear model developed in the study may be used to predict in-use fuel consumption based on data publicly available. The model utilizes the fuel consumption measured in type-approval, the mass, and the engine capacity to provide in-use fuel consumption. This may be either used to correct fuel consumption factors currently utilized by emission models (e.g. COPERT, HBEFA, VERSIT+, and others) or could be used independently to make projections on how fuel consumption may develop on the basis of changing future passenger cars characteristics.