Understanding population history and genetic connectivity is important for assessing the long-term viability of populations. The Atlantic surfclam (Spisula solidissima) has historically been subdivided into two subspecies: lesser-known S.s. similis in the South Atlantic Bight and Gulf of Mexico, and commercially important S.s. solidissima mostly north of Cape Hatteras. Novel populations of the “southern” S.s. similis subspecies were previously identified far north of their historical range limit. Here, to test a hypothesis of recency and isolation in the North, population samples from Southern New England S.s. similis were compared for the first time to a population sample from Georgia, as well as to S.s. solidissima. Population structure and demographic history were inferred using genetic variation at a combination of microsatellite, mtDNA, and nuclear intron loci. The Southern New England populations of S.s. similis had slightly lower nuclear DNA diversity than in Georgia. Genetic differentiation between Massachusetts and Georgia S.s. similis was weak and only significant for microsatellite markers. Consistent with weak differentiation, coalescent modeling of nuclear sequence variation indicated high levels of gene flow. All S.s. similis populations had extremely low mtDNA diversity, with only two mitochondrial cytochrome oxidase I haplotypes found in contrast to 15 haplotypes in S.s. solidissima. Collectively, the population genetic patterns are more parsimoniously explained by the northern S.s. similis populations originating from a postglacial expansion rather than recent colonization.