Climate division normals are calculated generally as combinations of values from stations that fall into each climate division. If the available stations do not exactly represent the entire climate division as a whole, biases in the normals will occur. This is a large issue in the western U.S., where stations are located primarily at low elevations. The result can be a negative biases in precipitation and positive biases in temperature. In addition, stations available for calculating divisional statistics change over time, which can produce temporal biases. A more stable method in both space and time for calculating divisional statistics is to derive them from gridded data sets that account for spatial variations between stations. These include the effects of elevation, rain shadows, coastlines, and others. These factors are relatively stable in time, and render the areal averages much less susceptible to variations in station availability. One of these gridded data sets is the peer-reviewed PRISM data set of 1961-90 mean annual precipitation, produced by Oregon State University’s Spatial Climate Analysis Service and supported by the USDA Natural Resources Conservation Service. Grid cell resolution is 4 km. Using GIS and divisional polygon coverages, 1961-90 divisional averages were calculated from the PRISM data set and compared to the present divisional averages.