How urbanization shapes population genomic diversity and evolution of urban wildlife is largely unexplored. We investigated the impact of urbanization on white-footed mice, Peromyscus leucopus, in the New York City metropolitan area using coalescent-based simulations to infer demographic history from the site frequency spectrum. We assigned individuals to evolutionary clusters and then inferred recent divergence times, population size changes, and migration using genome-wide SNPs genotyped in 23 populations sampled along an urban-to-rural gradient. Both prehistorical climatic events and recent urbanization impacted these populations. Our modeling indicates that post-glacial sea level rise led to isolation of mainland and Long Island populations. These models also indicate that several urban parks represent recently-isolated P. leucopus populations, and the estimated divergence times for these populations are consistent with the history of urbanization in New York City.