There has been a recent surge in the literature outlining methodologies that make use of spatially extraneous yield data in estimating crop insurance premium rates. The idea of borrowing information across space to better estimate tail probabilities is appealing. Along a different vein, recent research has questioned the validity of using whatever limited historical yield data exists given the number of technological changes in seed and farm management technologies as well as climate change. This literature has suggested historical yield data be trimmed to the most recent 25-30 years, thereby making the historically discarded yield data temporally extraneous. In this manuscript, we present three successively flexible data-driven methodologies to nonparametrically smooth across both space and time simultaneously. We apply these methodologies in estimating U.S. corn and soybean county-level crop insurance premium rates. Wefind significant borrowing of information across both time and space. We also find all three methodologies improve both the stability and accuracy of crop insurance premium rates.