Spatially-resolved climate field reconstructions are ideal for analyzing spatial anomaly patterns and characterizing regional-scale trends resultant from climate change. To date, few fine-scale, spatially-resolved paleotemperature datasets exist in the Northern Hemisphere. Here, we present a 2.5x2.5o temperature field reconstruction of warm season (April-August) mean surface air temperatures, developed from a network of 130 tree-ring chronologies. In the reconstruction’s current form, statistical calibration and validation tests indicate that tree-ring predictors for each of the individual grid points provide robust, multi-centennial to multi-millennial length estimates of historical temperature variability across many parts of North America. Further, we identify areas of North America where more data are needed to improve both the spatial and temporal gaps in the coverage of the paleotemperature atlas. A completed paleotemperature atlas will allow us to examine and compare the historical presence, persistence, and modes of external forcings of trends in surface air temperatures, and it will improve our understanding of the spatiotemporal relationships between temperature and other climatic variables across North America. We highlight the importance of increasing the density of temperature-sensitive tree-ring predictors in North America by the creation of new collections, but we also strongly emphasize the utility of re-examining preexisting tree-ring collections using novel techniques such as blue intensity. As such, we propose network-wide collaboration of tree-ring researchers in order to improve this dataset for its optimum effectiveness and usage by the broader paleoclimate community.