This paper reviews in-use emissions from over-the-road heavy-duty (greater than 8500 lb gross vehicle weight) diesel vehicles. Emissions of particulate matter (PM), oxides of nitrogen (NOx), carbon monoxide (CO), and hydrocarbon (HC) are considered. Emissions measurements have been made by chassis dynamometer, by remote sensing, and in tunnel studies. There is reasonable agreement between the different types of studies for emissions of NOx, but agreement for emissions of other pollutants is qualitative. Remote sensing studies indicate that emissions of NO are normally distributed, while emissions of CO and HC are skewed to a few high emitting vehicles. Chassis dynamometer results indicate that average emissions of PM, CO, and HC have been reduced during the past two decades, but average emissions of NOx have not changed. Thus, emissions regulations for PM have been somewhat effective, although the degree of PM reduction is less than expected based on changes in the standards. Emission regulations have apparently not been effective at reducing in-use NOx. Chassis dynamometer emissions of PM are widely variable, typically spanning an order of magnitude for any given model year. This variability is attributed to variation in testing altitude and temperature, inertial weight, test cycle, vehicle condition (deterioration), and driver variability. The fleet of vehicles tested by chassis dynamometer has been compared to the in-use fleet and found to overemphasize transit buses (over one-third of the tested fleet). In comparison, there are estimates that transit buses generate less than 5% of the vehicle miles traveled for heavy-duty vehicles. The tested fleet was also found to be newer and of lower mileage than the in-use fleet. Several improvements to study methodology and new research directions are proposed.