European Type approval procedure defines a synthetic driving cycle (the NEDC) over which one vehicle per type has to be tested. Euro 1, 2, 3, 4 and 5 differ (beside vehicle preconditioning and warm-up procedures introduced since Euro 3) only because limits for the different pollutants have been progressively lowered.This paper analyses through a number of experimental tests on spark-ignition cars, a hybrid and a conventional vehicle, the driving conditions responsible for most of the emissions and assesses how such conditions are reproduced by the type approval test.The engine conditions mostly responsible for emissions are: warm-up phase, full loads and transients. Only the warm-up is well covered by the NEDC for vehicles with more than 35 kW/ton power-weight ratio.Tests performed with the Honda Hybrid (the low environmental impact vehicle) and with the Alfa Romeo 147 1.6 (the conventional vehicle) showed how on the NEDC most emissions are produced in the warm up phase and are of the same order of magnitude for both cars while on more realistic driving cycles (ARTEMIS cycles have been used here) the capacity of the Hybrid to mitigate the transients effects results in a much lower emission rate. Full load conditions are not kept under control by the Hybrid as by the conventional vehicle when the O2 exhaust sensor is disabled and the air-fuel ratio is no more stoichiometric.Even though this paper has not tested all possible vehicles types, it shows that European type approval procedure has some weaknesses in accounting for the main causes of vehicle emissions. Any new procedure addressing better transients and full load conditions would help the diffusion of low emission vehicles like hybrids more than progressively lowering the allowed emission thresholds on current type approval procedure.