Electric vehicles have a unique potential to transform personal transportation. We analyze this transition with a dynamic model capturing falling costs of electric vehicles, decreasing pollution from electricity, and increasing vehicle substitutability. Our calibration to the US market shows a transition from gasoline vehicles is not optimal at current substitutability: a gasoline vehicle production ban would have large deadweight loss. At higher substitutability, a ban can reduce deadweight loss from vehicle mix and adoption timing inefficiencies. A cumulative gasoline vehicle production quota has smaller deadweight loss, and an electric vehicle purchase subsidy is more robust to regulator misperceptions about substitutability.