Simulations from the Canadian Coupled General Circulation Model version 3.1 Special Report on Emissions Scenarios A2 and A1B and the artificial neural networks (RNA) for downscaling statistically the maximum temperature and the minimum temperature daily values to the Alberni Robertson Creek weather station level, located at the Vancouver Island, Canada were used. The data generated for the station was analyzed and mean and variance were estimated; in addition, comparisons between the values for each scenario in the base period (1961-2000) and the simulations in the 21st century were carried out. The results show an increase in the values of the minimum and maximum temperature means between 1.16 and 1.47 Celsius degrees in the zone for the 21st century. The models developed accurately simulated the temperature inter-annual cycles, as well as the series mean temperature. However, the variance of the original series is greater than that of the model for the period recorded. The method employed proved to be flexible and easy to implement, with low computational requirements. Given these characteristics, using it in other regions which have reliable records for macro climatic variables is recommended.