Plug-in electric vehicles (PEVs) can contribute to deep greenhouse gas (GHG) reduction targets but their efficacy depends on the sources of electricity. PEV GHG intensity can vary over time (and regionally), making it unclear how policymakers should regulate PEVs in the short and long-term. To explore this uncertainty, the authors model the short-term (Study 1) and long-term (Study 2) well-to-wheels GHG intensity of PEVs in three regions with very different electricity grid profiles: the Canadian provinces of British Columbia, Alberta, and Ontario. Study 1 uses empirical data on vehicle preferences, driving patterns, and recharge access from a representative survey of new vehicle buyers in Canada (n = 1754) to construct a temporally-explicit model of PEV usage in 2015. Fleet-wide emissions intensity of PEVs varies substantially between regions, with the greatest reduction potential relative to conventional gasoline vehicles seen in British Columbia (78–98%), followed by Ontario (58–92%) and Alberta (34–41%). Study 2 simulates the potential long-term dynamics of technology, behavior, and emissions with the CIMS energy-economy model. With the emissions intensity of electricity decreasing by at least one-third by 2050 and vehicle energy efficiency improving over time, simulation results find that, compared to 2015, 2050 fleet average PEV emissions are 40–52% lower in British Columbia, 57–74% lower in Alberta, and 36–46% lower in Ontario. Overall, these studies find that PEVs offer substantial GHG emissions benefits compared to conventional vehicles in all scenarios explored. Policy makers seeking deep GHG cuts may want to support PEV adoption, even in jurisdictions that presently use relatively carbon-intensive electricity.