Evaporative emission levels have been determined in a CONCAWE* programme for a range of ten uncontrolled European vehicles using a modified SHED test procedure as developed by the CEC*. Three extra vehicles were tested which were equipped with evaporative and exhaust emission control systems, but of the same make and model as three of the uncontrolled test cars.The vehicles were tested using several warm-up cycles and on a range of fuels whose volatility parameters were independently varied, including oxygenate blends. Exhaust emissions were determined and a few measurements of true diurnal emissions carried out.Vehicle fuel system design had the greatest effect on evaporative emissions which varied between 4 to 16 g/test on a typical European summer fuel. Gasoline volatility had a significant but smaller effect and RVP was shown to be the dominant fuel parameter. At the same volatility, oxygenate blends gave similar or lower emissions than hydrocarbon fuels.Hot-soak and running losses increased significantly with increasing warm-up cycle severity. True diurnal emissions were found to be significant and of similar magnitude to combined hot-soak and running losses.The carbon canister emission control systems tested were very effective and reduced emissions by up to 85 per cent.