In marine environments, many species have apparently colonized high latitude regions following the last glacial maximum (LGM) yet lack a life-history stage, such as a free-living larva, that is clearly capable of long-distance dispersal. Two hypotheses can explain the modern high latitude distributions of these marine taxa: (1) survival in northern refugia during the LGM or (2) rapid post-glacial dispersal by nonlarval stages. To distinguish these two scenarios, I characterized the genetic structure of two closely related northeastern Pacific gastropods that lack planktonic larvae but which have distributions extending more than 1000 km north of the southern limit of glaciers at the LGM. Despite having identical larval dispersal potential, these closely related species exhibit fundamentally different patterns of genetic structure. In Nucella ostrina, haplotype diversity among northern populations (British Columbia and Alaska) is low, no pattern of isolation by distance exists and a coalescent-based model of population growth indicates that during the LGM population size was reduced to less than 35% of its current size. In the congeneric and often sympatric N. lamellosa, northern populations harbour a diversity of ancient private haplotypes, significant evidence of isolation by distance exists and regional subdivision was found between northern (Alaska) and southern (southern British Columbia, Washington and Oregon) populations. Estimates of coalescent parameters indicate only a modest reduction in population size during the LGM and that northern and southern populations of N. lamellosa split ∼50 Kyr before the LGM. The patterns are consistent with the hypothesis that N. ostrina recently reinvaded the northeastern Pacific but N. lamellosa survived the LGM in a northern refuge. A comparison of similar studies in this region indicates that depleted levels of genetic variation at high latitudes — evidence suggestive of recent colonization from a southern refuge — is more common among intertidal species that live relatively high on the shore, where exposure times to cold stress in air are longer than for species living lower on the shore. These data suggest that for some faunas, ecological differences between taxa may be more important than larval dispersal potential in determining species’ long-term biogeographical responses to climate change.