Dispersal among populations is essential to ensuring the persistence of marine species, particularly those which are targeted in our fisheries. The dispersal potential, or connectivity, of most species is largely determined by the early life history characteristics and the geographic setting of populations. Here, we investigated metapopulation persistence using a network approach to analyse and visualise metapopulations dynamics. Traditional demographic approach for studying population dynamics has been implemented to explicitly include dispersal dynamics to account for losses and gains among subpopulations. We focused on two important fish species, the snapper (Chrysophrys auratus, formerly Pagrus auratus) and the King George whiting (Sillaginodes punctatus), found throughout the South-East Marine Region of Australia. These species represent contrasting life history traits, the snapper is characterised by a short pelagic larval duration phase, whereas the King George whiting larvae spend a much longer time in the larval stage. We used a biophysical larval dispersal model to simulate dispersal behaviour and develop metapopulation networks, where habitat patches with subpopulations are represented by nodes and dispersal connections are represented by linkages. We calculated and identified network metrics indicative of persistence to demonstrate the power of this approach in identifying key drivers of metapopulation persistence.