There are many methods applied to spatially discontinuous station records to produce gridded data and the Daymet and PRISM methods are in wide use. Spatially and temporally continuous meteorological data is an important input in hydrological and ecological modeling, yet the level of uncertainty associated with the interpolated data is largely unknown. This study used the annual mean, annual maximum, and annual minimum temperatures from 1980-2003 from 8,126 ground stations as well as the Daymet and PRISM models. These three parameters allowed us to quantify the ability of PRISM and Daymet to reproduce measured temperatures over the coterminous United States. The goal of this project is to quantify the ability of each model to reproduce temperature for the study period over the entire coterminous United States, including evaluation by geographic region and elevation zone. We filter the data using a quality control process and then apply the logs odds ratio (LOR), mean squared error, and mean signed error to asses the performance of the two methods. For all three temperature parameters results demonstrated a difference in performance depending on location and elevation, but overall PRISM performs better then Daymet.