Spatial distribution of species and thus the composition of deciduous forests are expected to change in response to global climatic warming. Climate-growth response algorithms in forest simulation models that represent the influence of climate on tree growth are based on the geographic range limits of a species. These models assume that temperature limitations control the latitudinal range boundaries for tree species in the eastern United States, and that water limitations control the western range boundary. This study investigated spatial variation of tree-growth responses to historical variations in climate using tree-ring data for Quercus rubra L. from 71 study sites across latitudinal and longitudinal temperature and precipitation gradients in eastern North America. Correlation analyses were used to identify significant site-specific associations between radial growth indices and 130 climate variables for the common time interval 1930-1980. Geographic information systems (GIS), and Mantel and partial Mantel spatial correlation analyses were used to map, identify, and measure spatial associations between oak sensitivity to climate and regional climate patterns to test the hypothesis that climate controls the spatial distributions of species range limits.Seasonal climate variables exhibited the strongest correlations with radial growth most consistently across the 71 study sites. However, spatial patterns in regional climate were only weakly associated with spatial variation in red oak sensitivity to climate variables. Contrary to expectations, correlations between red oak radial growth indices and temperature variables at northern and southern sites were not significant, failing to support the hypothesis that latitudinal range limits are defined by temperature effects on mature tree growth. High early growing season temperature reduced growth mostly at sites in the southwest region, which concomitantly experienced high inter-annual variability in soil moisture. Increased early growing season site water balance and precipitation was associated with greater radial growth in the western third of red oak's range, partially supporting the hypothesis that the western range limit is defined by limited water availability. These results indicate that climate only partially determines range limit locations; thus model projections may exaggerate forest responses to climate change. / Department of Biology