During the late Eocene in southern Australia, biosiliceous spiculite and spongolite accumulated within shallow marine estuaries and embayments inboard of an open-ocean, low-energy, cool-water carbonate shelf. Estuarine and embayment waters were ideal for an otherwise deep-water sponge assemblage to colonize these paleoenvironments and outcompete all other organisms. Shallow, warm-water sponge growth was promoted by (1) a flat, humid, rainforested terrestrial hinterland with little surface runoff and high groundwater flow rates that led to an enhanced export of dissolved silica, low sedimentation rates, and oligotrophic inshore environments; (2) a paleogeography that protected silica-rich estuarine and embayment waters from dilution by ocean waters, and helped maintain calm conditions; (3) low siliciclastic sedimentation rates that enhanced substrate ventilation by frequent bioturbation, occasional storms, and diffusion; (4) muddy substrates and occasional turbidity that inhibited growth of phototrophic and calcareous sedentary epibenthic organisms; (5) substrates that lacked carbonate or organic matter (due to low productivity) and were consequently suboxic with high pH, a situation that drove extensive iron oxide-biogenic opal-kaolinite dissolution and clinoptilolite-verdine precipitation immediately below the sediment-water interface. Substrate ventilation helped cycle dissolved silica between substrate pore waters and overlying marine waters where the silica was fixed in epibenthic sponges. This silica cycle, combined with export of terrestrial silica, maintained high dissolved silica concentrations within embayments and estuaries. The result was late Eocene embayments-estuaries along 2000 km of the southern Australian margin that were both chemically and physically ideal for sponge colonization. The general controlling paleoecological variables of these late Eocene environments are applicable to both modern deep-sea sponge habitats and ancient shallow-water, biosiliceous sponge facies.