Adaptation and gene flow are insufficient to rescue a montane plant under climate change.

Climate change increasingly drives local population dynamics, shifts geographic distributions, and threatens persistence. Gene flow and rapid adaptation could rescue declining populations yet are seldom integrated into forecasts. We modeled eco-evolutionary dynamics under preindustrial, contemporary, and projected climates using up to 9 years of fitness data from 102,272 transplants (115 source populations) of Boechera stricta in five common gardens. Climate change endangers locally adapted populations and reduces genotypic variation in long-term population growth rate, suggesting limited adaptive potential. Upslope migration could stabilize high-elevation populations and preserve low-elevation ecotypes, but unassisted gene flow modeled with genomic data is too spatially restricted. Species distribution models failed to capture current dynamics and likely overestimate persistence under intermediate emissions scenarios, highlighting the importance of modeling evolutionary processes.