Range expansions, whether they are biological invasions or climate change-mediated range shifts, may have profound ecological and evolutionary consequences for plant-soil interactions. Range-expanding plants encounter soil biota with which they have a limited coevolutionary history, especially when introduced to a new continent. Past studies have found mixed results on whether plants experience positive or negative soil feedback interactions in their novel range, and these effects often change over time. One important theoretical explanation is that plants locally adapt to the soil pathogens and mutualists in their novel range. We tested this hypothesis in Dittrichia graveolens, an annual plant that is both expanding its European native range, initially coinciding with climate warming, and rapidly invading California, after human introduction. In parallel greenhouse experiments on both continents, we used plant genotypes and soils from five locations at the core and edge of each range to compare plant growth in soil from D. graveolens populations and nearby control microsites as a measure of plant-soil feedback. Plant-soil interactions were highly idiosyncratic across sites in each range. On average, plant-soil feedbacks were more positive in the native range than in the exotic range. In line with the strongly heterogeneous pattern of soil responses along our biogeographic gradients, we found no evidence for evolutionary differentiation between plant genotypes from the core to the edge of either range. Our results suggest that the evolution of plant-soil interactions during range expansion may be more strongly driven by local evolutionary dynamics varying across the range than by large-scale biogeographic shifts.