Motor gasoline and diesel contribute 30% to total energy related carbon dioxide (CO2) emissions in the U.S. However, this estimate only accounts for emissions from direct combustion and does not include indirect emissions from processing and fuel movement, even though indirect (scope 3) CO2 emissions are a significant contributor. Gasoline and diesel flow through a complex supply chain from oil extraction to point of combustion and estimates of their indirect emissions are typically aggregated as national or regional averages and not available at county or city scale. This dissertation presents a novel method to quantify U.S. supply-chain CO2 emissions to the county-scale for gasoline and diesel consumed in the on-road sector. It considers how these fuels flow across the U.S. petroleum infrastructure consisting of pipelines, tankers, trucks, trains, refineries, and blenders. It resolves county-scale indirect CO2 emissions using publicly accessible data to allocate fuel movement between different links and transportation modes across the country. For most of the U.S., the exact volume of fuel moved between counties from combinations of refineries and transportation modes is not explicitly known. To estimate these fuel movements, I use linear optimization with supply and demand related constraints. Estimating on-road gasoline and diesel indirect CO2 emissions at high spatial resolution finds that on-road gasoline CO2 emissions increase by 24% and on-road diesel CO2 emissions increase by 18%. For both fuels there are large variations in the carbon intensity (kgCO2/gal) across the country and the relationship of county carbon intensity with explanatory variables related to fuel supply infrastructure is tested. Regression results indicate that presence of interstate highways, refineries and blenders are inversely related to carbon intensity while presence of fuel pipelines increases diesel carbon intensity. Finally, the on-road gasoline scope 3 CO2 emissions results are assessed in relation to indirect CO2 emissions from electricity consumption at the county scale to analyze the effectiveness of future electric vehicle (EV) transition actions. In this analysis, states with existing EV transition mandates (zero emission vehicle or ‘ZEV’ states) are shown to have on average 12% higher CO2 emissions reduction when transitioning to EVs, over non-ZEV states.