Local adaptation is a critical evolutionary process that allows plants to grow better in their local compared to nonnative habitat and results in species-wide geographic patterns of adaptive genetic variation. For forest tree species with a long generation time, this spatial genetic heterogeneity can shape the ability of trees to respond to rapid climate change. Here, we identify genomic variation that may confer local environmental adaptations and then predict the extent of adaptive mismatch under future climate as a tool for forest restoration or management of the widely distributed high-elevation oak species Quercus rugosa in Mexico. Using genotyping-by-sequencing, we identified 5354 single-nucleotide polymorphisms (SNPs) genotyped from 103 individuals across 17 sites in the Trans-Mexican Volcanic Belt, and, after controlling for neutral genetic structure, we detected 74 FST-outlier SNPs and 97 SNPs associated with climate variation. Then, we deployed a nonlinear multivariate model, Gradient Forest (GF), to map turnover in allele frequencies along environmental gradients and predict areas most sensitive to climate change. We found that spatial patterns of genetic variation were most strongly associated with precipitation seasonality and geographical distance. We identified regions of contemporary genetic and climatic similarities, and predicted regions where future populations of Q. rugosa might be at risk due to high expected rate of climate change. Our findings provide preliminary details for future management strategies of Q. rugosa in Mexico and also illustrate how a landscape genomic approach can provide a useful tool for conservation and resource management strategies.