The objective of this project was to assess the effects of various blends of biodiesel on locomotive engine exhaust emissions. Systematic, credible, and carefully designed and executed locomotive fuel effect studies produce statistically significant conclusions are very scarce, and only cover a very limited number of locomotive models. Most locomotive biodiesel work has been limited to cursory demonstration programs. Of primary concern to railroads and regulators is understanding any exhaust emission associated with biodiesel use, especially NOX emissions.In this study, emissions tests were conducted on two locomotive models, a Tier 2 EMD SD70ACe and a Tier 1+ GE Dash9-44CW with two baseline fuels, conventional EPA ASTM No. 2-D S15 (commonly referred to as ultra-low sulfur diesel - ULSD) certification diesel fuel, and commercially available California Air Resource Board (CARB) ULSD fuel. A single batch of soy-based B100 was blended with the EPA and CARB diesel fuels to yield 5 percent and 20 percent by volume blends of fuels. A randomized test matrix was used to perform triplicate tests on each of the six test fuels (EPA0, CARB0, EPA5, CARB5, EPA20, and CARB20).The results of these emissions test results were analyzed to determine the statistical relevance of any difference in emissions among fuels. General emissions and fuel economy trends for biodiesel seen in other studies and in other applications were seen in this study. Higher blend levels of biodiesel were associated with lower carbon monoxide and particulate matter, and higher levels of nitrogen oxides and fuel consumption. Diesel fuel with 20 percent biodiesel often resulted in statistically significant differences from the fuel with 0 percent or 5 percent biodiesel. The difference between 0 percent and 5 percent biodiesel was generally not statistically significant. Different trends between the locomotives could be explained by differences in emissions certification levels, combustion cycle (4-stroke vs. 2-stroke), and lubricating oil consumption.