Ecosystem-based fisheries managers are increasingly seeking quantitative and spatially-explicit information on species distributions to assist with the management of fisheries and aquatic habitats. In this study, we used boosted regression trees (BRT) to build species distribution models for a highly valued coastal teleost – pink snapper (Sparidae: Chrysophrys auratus) across rocky reefs adjacent to Australia's most urbanised coastline. BRT models for juvenile (<25 cm total length) and adult (>32 cm total length) snapper were created using a suite of environmental and habitat predictors. A surrogate for multiple anthropogenic stressors, measured as surrounding human population density, was also included in the models. The BRT model for juvenile snapper performed well (cross-validated AUC = 0.78) and identified habitat features as the most important drivers of their distribution across the region. Juvenile snapper were commonly associated with small patch reefs of low relief adjacent to large estuarine water bodies. In contrast, the performance of the BRT model for adult snapper was weak (cross-validated AUC = 0.68) but identified human population density over habitat features as the strongest predictor of adult snapper distributions. Lower occurrences of adult snapper were associated with reef habitats adjacent to large metropolitan centres, suggesting anthropogenic stressors, such as water pollution, noise and fishing may be negatively impacting adult snapper in the region. Our results highlight essential habitats for snapper populations, notably the importance of large estuaries in the coastal seascape, which are nurseries for juvenile snapper. Knowledge of the demographic habitat associations and spatial distribution of snapper across this highly urbanised coastline will support ongoing management and monitoring of snapper populations and their key habitats.