To study the impact of incremental climatic warming on summer extreme temperature event frequency, the historical record of daily maximum June, July, and August temperatures was analyzed for nine sites across Canada. It was found that all of these sites are well modeled by a first-order autoregressive process using three parameters: the mean, variance, and first-order autocorrelation coefficient. For slight changes in the mean or variance there are increases in the frequency of both single days and runs of 2–5 consecutive days with daily maximum temperatures over a threshold value. For example, for a 3°C increase in the mean daily maximum temperature at Toronto, the frequency of a 5-day consecutive run over 30°C rose by over a factor of 8 to 7.1%. Sites with less variability are more sensitive to an increase in the mean summer temperature than sites with higher variability. Analysis of simulated series indicates that when two parameter values change simultaneously the change in the frequency of a given event is usually greater than the sum of the individual changes. Output from the Canadian Climate Centre GCMII model for the nine sites for both the current and 2 × CO2 atmosphere indicate an average increase in the daily maximum temperature of 4.2°C. Changes in the standard deviation and autocorrelation were usually less pronounced. For Toronto, a positive correlation (R2 = 0.718) between daily peak power demand and the cube of the current and previous 2 days daily maximum temperature was found. A sensitivity analysis was performed on daily peak power demand by first generating temperature time series and then using the derived regression relationship. Results follow a predictable pattern and indicate that the standard deviation of the peak power series increases proportionally more than the mean for increases in the mean daily maximum temperature. For example, for a 3°C increase in mean daily maximum temperature, the increase in mean peak power demand was 7% (1200 MW) while the increase in the standard deviation of peak power demand was 22%. Changes in the autocorrelation of the temperature time series do not lead to significant changes in the mean or standard deviation of daily peak power demand. These results indicate that, while the average peak power demand is not moved drastically, the number of high energy consumption days may increase appreciably due to higher variability, placing stress on the provincial power utility to meet this higher demand.