Abstract
This presentation gives an account of the reliability of vehicle emission data. It includes the individual car as well as the entire population of vehicles. Two main sources of error exist. One is the problem of measurement, whether one uses data from a laboratory or data taken from measurements of real traffic. The second problem concerns the mathematical model used. The paper continues with two examples: (1) a calculation of exhaust fumes in Sweden until 2015, and (2) the content of emissions in a street. The examples are followed by a discussion of how the problems with uncertainty have been tackled.