A regional meteorology–chemistry model was used to assess the effects of passenger car conversion to battery electric vehicles (BEV) on summer O3 concentrations in Kanto (Japan’s most populous region). Four sensitivity experiments were conducted on different on-road and upstream (power plant and gas station) emission conditions. Daytime 8-h maximum O3 decreased by 3 ppb (5%) and 4 ppb (5%) in urban and inland suburbs, respectively. O3 levels decreased even in urban (VOC-limited regions) because exhaust and evaporative VOC emissions from vehicle and gas stations were reduced effectively (especially alkenes from gasoline evaporation; highly reactive in O3 formation). In the suburbs (NOx-limited regions), reduction of exhaust NOx by BEV shifting was significant, but in urban, even only evaporation measures induced almost the same O3 reduction effect as BEV shifting. The additional emissions from thermal power plants due to BEV night charging contributed little to the next day’s daytime O3 on a monthly average basis. However, on some days, pollutants were stored in the upper part of the stable nighttime boundary layer and could affect the surface O3 as the next day’s mixed layer development. Depending on the O3 sensitivity regime (NOx- or VOC-limited), additional NOx plumes from rural (urban) power plants tended to increase (decrease) the next day’s O3. However, the distribution of the regime changes temporally and spatially. The H2O2/HNO3 ratio was discovered to be a clear indicator for distinguishing regime boundaries and was effective in predicting positive or negative O3 sensitivity to the additional emissions from power plants.