For many benthic marine taxa, dispersal over large distances is dependent on a pelagic larval phase. It is through the dispersal process that benthic taxa can found new populations, colonise disturbed or degraded habitats and achieve genetic exchange between geographically separated subpopulations. While it is widely accepted that the dispersal process is important in determining the dynamics of many marine communities, difficulties associated with the direct measurement of larval dispersal mean that detailed knowledge of the mechanisms controlling larval transport has been elusive. Consequently, indirect methods, such as population genetics, have been used to estimate dispersal pathways. This thesis examines the population genetic structure of New Zealand’s coastal benthos with the aim of enhancing our understanding of population connectivity, as well as identifying physical and biological processes that might influence population genetic structure. The population genetic structure of New Zealand’s coastal benthos was examined in a quantitative literature review. Divergence between northern and southern populations was identified as the most frequently reported population genetic structure, with the divergence most often located in central New Zealand. Additional population structure although rare, was reported for a number of taxa, particularly those restricted to estuaries and with life history traits indicative of limited dispersal potential. A significant negative correlation between pelagic larval duration (PLD) and genetic divergence suggests that PLD may be a useful proxy for a species’ dispersal ability. However, variation in estimates of divergence for taxa with limited PLDs suggests that other factors may also influence dispersal potential. To determine whether populations of an estuarine organism with life history traits indicative of widespread dispersal are genetically subdivided, I examined the population genetic structure of the clam Austrovenus stutchburyi using the mitochondrial gene cytochrome oxidase c subunit I (COI). Analyses indicated that dispersal was limited and that gene flow was mostly occurring among estuaries in close proximity. Genetic boundaries were detected in central New Zealand, about the East and North Capes and in the south of the South Island. Similar boundaries have been reported for estuarine taxa lacking a dispersive larval phase suggesting that distribution and habitat requirements may influence patterns of gene flow. To further investigate gene flow among estuaries, I conducted a multi-scale genetic analysis of A. stutchburyi populations on New Zealand’s west coast Analyses of COI and microsatellite loci showed populations to be well connected within estuaries and at the regional sampling scale (≤ 156 km) although not at a national scale (≤ 1226 km). Distance among populations explained much of the observed genetic variation as did distance between estuaries. Long stretches of open coast appear to act as barriers to dispersal among A. stutchburyi populations. However, as long stretches of open coast also coincide with putative hydrodynamic dispersal barriers there is some uncertainty about the roles of barriers versus inter-estuary distance in generating the observed divergences. Within-estuary genetic differences were evident from COI analyses, but may result from local adaptation to within-estuary environmental gradients. I conclude that genetic connectivity among populations of New Zealand’s estuarine taxa is generally low relative to coastal species. The data presented here also suggest that migration among patchily distributed habitats, such as estuaries, will be dependent on the spatial arrangement of habitats. Consequently, patchily distributed taxa may experience low rates of ecologically meaningful connectivity, requiring relatively cautious management at small spatial scales to ensure the persistence of intact biological communities in the face of persistent anthropogenic and natural disturbances.