Climatic and geological changes have been recognised as fundamental mechanisms in the evolution of populations. This thesis assesses population genetic structures and dispersal patterns of arthropods in two geographic regions-Antarctica and New Zealand-using molecular techniques in conjunction with distributional data and dispersal experiments. Regional genetic divergence was revealed using allozymes and mtDNA (COI) for populations of the endemic Antarctic collembolan Gomphiocephalus hodgsoni from southern Victoria Land and Ross Island (Ross Dependency). Genetic discontinuities across geographical barriers (e.g. sea-ice) suggest limited dispersal opportunities and long-term habitat fragmentation throughout the Pleistocene. In addition, the identification of two sympatric and genetically divergent groups throughout one of the continental sites (Taylor Valley) suggests that these mechanisms have been conducive to speciation. The limited mixing of mtDNA haplotypes between one island site (Cape Bird) and one continental site (Granite Harbour) is unlikely to be accounted for by the un-aided dispersal capacity of G. hodgsoni and recent human- or bird-mediated dispersal is highly probable. Furthermore, recent localised dispersal of Collembola was identified in two continental sites (Taylor Valley and Granite Harbour) through comparisons with previous distributional studies, and suggests that range expansion of up to 5 km has occurred within the last 40 years. The endemic estuarine amphipods Paracorophium lucasi and P. excavatum from North, South and Chatham Islands of New Zealand were examined using allozyme analyses that also identified clear genetic breaks across geographical barriers (e.g. land-bridges) separating biogeographic regions. However, populations of P. lucasi were more divergent than populations of P. excavatum over similar geographic distances, but in most cases gene flow appears to maintain a homogenous population genetic structure in populations that share a common coastline. These results are congruent with a high rate of female-biased juvenile dispersal that was identified during field experiments with Paracorophium spp. in Tauranga Harbour. Such dispersal may be a mechanism to avoid inbreeding and inter- and intraspecific competition. During these experiments, I identified the New Zealand Paracorophium species, as well as P. brisbanensis, previously recorded only from Australia. I conclude that this latter species is unlikely to be indigenous to New Zealand or the result of natural dispersal from Australia, but rather anthropogenic translocation (i.e. shipping activities). This research reveals limited dispersal and high levels of genetic divergence for three arthropod taxa from fragmented habitats in Antarctica and New Zealand. I conclude that such levels of cryptic diversity indicate the inadequacies of morphology-based classification schemes. Accordingly, assessments based on genetic diversity (e.g. mtDNA analyses) are required.