Heavy duty diesel engines are homologated for emissions using engine test dynamometers according to predefined emission cycles. In 2012, it was released in the Brazilian market the first diesel heavy duty vehicles with PROCONVE P7 technology (EURO V). The homologation tests include emission measurement according to stationary cycles (ESC) and also transient tests (ETC). For the ESC emission measurement it is possible to use a passive dynamometer, while the transient test requires an active dynamometer. Uncertainty of measurement is an important matter to get accurate emission data, but there are not many published papers about a method for its calculation related to diesel engine emissions. According to this scenario, this paper presents a method to calculate the uncertainty of measurement of legislated pollutants for diesel engines (CO, NOx, THC and particulate matter). The experiments were done using a passive dynamometer and an ESC cycle, with exhaust emissions being measured by a gaseous emission bench. Complete ESC tests were repeated on a EURO III diesel engine, allowing type A uncertainty calculation. Also type B uncertainty (due to equipment calibration, resolution etc.) was studied and latter calculated. This paper will report the calculation method applied to estimate the different uncertainty budgets of a diesel engine emission test and also some results of uncertainties of measurement. It will also present some discussions about the most important variables that contribute for the final uncertainty values of the emission test.