Bacterial wilt (BW) caused by Ralstonia solanacearum (Rs) is one of the most harmful plant disease in the world1 • Special careful should be paid to phylotype I due to its large host range, its worldwide distribution, and its high evolutionary potentiai2. However there is a Jack of studies focusing on population genetics of Rs, whereas basic knowledge on the dynarnics of pathogen genetic adaptation is now needed to predict plant resistance durability3, especially at the field scale. Thanks to a new ML V A scheme (ML V A-8) specifically tuned for field-scale microevolution of Rs phylotype :c4, we addressed the temporal and spatial dynamics of Rs populations' genetic structure in a field submitted to eggplant selective pressure. We set up a three years experimental field containing 3 microplots of resistant (E6) and susceptible eggplant (E8), two crops cycle were implanted each year (Cl to C5). No bacterial wilting was observed on resistant eggplant (E6) and we only isolated strains from latent infections on E6 during Cl, but no trade-offfor colonization of resistant eggplant was observed. On E8, we monitored an increase of incidence between Cl and the other cycle but virulence assay did not show a gain of virulence ofstrains between Cl and C2. Across eggplant cycles, the disease seemed to spread by host-to-host dissemination as well as by hot-spots of disease which finally merged. A clear genetic diversity reduction was observed in response to repetitive crop cycle of eggplant associated with the increased of frequency of two main haplotypes. However we failed to identify an effect of the host variety, the host's genetic background or the habitat of origin (soil, rhizosphere and stem) on the genetic structure of Rs populations. Our study helps to decipher the forces goveming of evolution of a field Rs population in response to host selective pressure, and the main outputs will be further discussed in the context of durable control ofBW in the field.