While electric vehicles including plug-in hybrid electric vehicles (PHEVs) and battery-powered electric vehicles (BEVs) are considered as promising alternative vehicle/fuel systems to significantly reduce petroleum consumption of the transportation sector, it is important to analyze the emission characteristics and to assess the emission reduction potentials of electric vehicles so that their environmental impacts in terms of climate change, air quality, as well as human health effects could be better understood. To fulfill this objective, we explicitly analyzed the emission characteristics of greenhouse gases (GHG) and criteria air pollutants (CAP, representing VOC, CO, NOx, PM₁₀ and PM₂.₅, and SOx,) of the U.S. power sector, a pivotal upstream sector that impacts the life-cycle GHG and CAP emissions associated with electric vehicles. In particular, we established statistically robust probability distribution functions (PDFs) to address the uncertainty and variation in both thermal and environmental performances of various types of power plants that can result in uncertainties of the life-cycle GHG and CAP emissions of electric vehicles. With the detailed characterization of the GHG and CAP emissions from the power sector, we performed a life-cycle analysis (LCA) of the GHG and CAP emissions of electric vehicles in the United States, with the GREET (Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation) model developed at Argonne National Laboratory. The LCA results with the inherent uncertainties investigated illuminate the real-world chances of GHG and CAP emission reduction benefits that are influenced by uncertain factors when introducing electric vehicles to displace gasoline vehicles in the United States.