Three heavy-duty vehicles (two buses and a truck) were tested on a chassis dynamometer and the engines were removed and tested on an engine dynamometer. Two 1993 DDC Series 50 and one 1993 Navistar DTA-466 were the engine models tested. The objective of this testing was to compare in-use emissions from vehicles with engine FTP emissions, and begin to evaluate and begin to evaluate the relationship between emissions on these two tests. Engine testing results (heavy-duty FTP) show emissions above certification levels for the two bus engines, especially for NOx, however the truck engine met the emissions standards. Chassis testing was performed using the Central Business District cycle and the EPA Urban Dynamometer Driving Schedule for Heavy-Duty Vehicles (heavy-duty transient truck or HDT cycle). There was a substantial difference in emissions for these two cycles on a g/mi basis, however both cycles gave essentially the same gaseous emissions on a g/gal of fuel consumed basis. Chassis test inertial weight had a significant effect on g/mi emissions of all pollutants, however on a g/gal basis, emissions were not a function of inertial weight except for PM, which was a weak function of inertial weight. The data were used to estimate bhp-h/mi factors for converting engine test data to the g/mi basis normally employed in pollutant inventory estimation. Overall the calculated factors were significantly different from those currently employed by EPA. When both engine test and chassis test data are converted to a g/gal basis, it was found that both tests produce approximately the same emissions. The use of fuel consumption based conversion factors therefore, may prove more accurate than the current approach.