Larval connectivity is critically important for the persistence of marine metapopulations, particularly considering accelerated climate change and growing fishing pressures. Now, more than ever, effective management plans are dependent on reliable connectivity data to help determine where and when to act. Unfortunately, for many marine species, population connectivity is not well known. Indeed, we often know very little about the larval life histories, spawning seasonality and habitat, and the relevant oceanography. Here, we demonstrate how a biophysical modelling approach and network analysis can be used to quantify the metapopulation implications of these uncertainties. Specifically, we estimate the spatial and temporal patterns in population connectivity across southern Australia and New Zealand resulting from long-distance dispersal in several decapod species (giant crab, giant spider crab, southern rock lobster). Results estimate broad-scale stock structure, highlight source-sink dynamics, identify important larval sources, and reveal the geographic structure of metapopulation connectivity across species. We show where and why uncertainty overwhelms predictions and how the modelling framework can be used to gain some valuable insights. Finally, we take a management perspective and suggest what results might be useful, and where they should be ignored.