Diesel locomotives provide versatility to the railway system as they do not depend on electric feeding. Worldwide they constitute important parts of locomotive fleets used for various tasks such as mainline services, shunting and terminal operations. The main drawbacks are emissions of CO2, NOx and other pollutants. The emissions to air are in general controlled by legislation. The European testing and approval procedure for locomotive engines is performed with the ISO 8178-F test cycle. It is alone meant to resemble all European locomotives. Good correlation was found between mainline operations and the ISO 8178-F test cycle. A good correlation was furthermore found between shunting operations and North American shunting test cycles. However, this thesis has shown that this results in inadequate description of for instance shunting operations. Typical characteristics for shunting are relatively high ratio of transient loading, low mean power output and high amount of idling. Therefore important aspects of diesel locomotive usage are thereby disregarded in certifications. Deficient correlation between test cycle and intended operations may lead to locomotives that are not properly optimised for the type of operation they are intended for. Improperly dimensioned and optimised locomotives cause impaired fuel economy and more emissions emitted. Many Swedish diesel locomotives are to a large extent used in shunting-like conditions in yards and industries for instance. Therefore, it is recommended to supplement the present ISO test cycle with a test cycle that includes the characteristic of shunting. Knowledge about duty cycles for specific operations can be of importance in locomotive acquisition, dimensioning of new locomotives or when estimating emissions and/or fuel consumption. It is essential to make proper dimensioning of propulsion systems as it can imply lower life cycle costs as well as fewer emissions. Appropriate dimensioning of the propulsion system is particularly important for non-conventional propulsion systems. If the application range of a test cycle is narrow, its emulating capability ought to be better at describing and representing the indented application. Today's test cycles state fractional power or torque as a function of engine speed. This may result in misleading duty cycles when the same kind of operation is undertaken with different locomotives that have widespread power ratings. A more powerful locomotive used for the very same load as a less powerful locomotive will result in a different duty cycle. Non-conventional propulsion systems are gaining popularity also in the railway industry where implementation of new technology usually is slow. Dual mode and hybrid systems are two examples. Related to test and duty cycles the build-up of these new propulsion systems change the prerequisites of the propulsion systems. From a duty cycle perspective it is reasonable to implement an additional test cycle for non-conventional propulsion systems if the popularity rises. / QC 20111115