Four sets of grids of normals (1961-1990 averages) of temperature maximum and minimum, and total precipitation were created for Canada. Evenly spaced, geo-referenced grids are better suited for certain purposes than original observations from the irregularly distributed stations. Grids provide better spatial coverage for regional climate change and variability studies; they can be easily integrated into Geographic Information Systems to build environmental models for forestry and agriculture, to study climate change scenarios and impacts, to calculate water budgets, etc. Fields of temperature and precipitation normals – the two primary climatological elements – constitute a very important reference field for this kind of scientific research. There are many interpolation techniques and approaches to gridding and mapping – in this case four academic and government institutions used the following schemes: - ‘square-grid’ technique based on multivariate regression model; - PRISM, or Parameter-elevation Regressions on Independent Slopes Model; - ANUSPLIN model based on partial thin plate smoothing splines; - Inverse Distance Weighting (IDW), a type of weighted average interpolator. Each approach is based on specific assumptions and often requires additional variables, such as elevation (Digital Elevation Model). Each method have certain known advantages and caveats and can produce different results, for example, some perform better in the mountainous or data sparse regions, some put a lot emphasis on topography alone. The purpose of this study is to summarize major characteristics of each method and ensuing grid sets, and then compare the grids in order to establish major biases and differences between them. This will be done for Western Canada, because PRISM grids cover only western provinces. This intercomparison should help users in making informed decision about the suitability of the grids for their purposes.