Exhaust emission tests run on an engine dynamometer are compared with those run on a chassis dynamometer. The worst case average difference between the chassis and engine dynamometer mass emissions, obtained over a period of 6 months, was less than 7%. The elimination of the driver, vehicle, and chassis dynamometer yielded significant improvement in test reproducibility with respect to carbon monoxide and oxides of nitrogen.A digital computer was used to control speed and throttle position of the engine dynamometer experiments. No transmission was used. The computer and engine dynamometer duplicated transient effects of transmission, vehicle, and chassis dynamometer.