-Fifty-six sites, representing 53 wetlands, were surveyed in the southern Carnarvon Basin in 1994 and 1995 with the aim of documenting the waterbird and aquatic invertebrate fauna of the region. Most sites were surveyed in both winter and summer, although some contained water only one occasion. Altogether 57 waterbird species were recorded, with 29 292 waterbirds of 25 species on Lake MacLeod in October 1994. River pools were shown to be relatively important for waterbirds, while many freshwater claypans were little used.At least 492 species of aquatic invertebrate were collected. The invertebrate fauna was characterized by the low frequency with which taxa occurred: a third of the species were collected at a single site on only one occasion. Patterns of occurrence were not strongly seasonal. Many undescribed species were found and many range extensions were. recorded, reflecting lack of previous aquatic invertebrate work in the region. The level of regional endemicity could not be assessed adequately, although it is probably comparatively low. In terms of their invertebrate fauna, five types of wetlands were distinguished: river pools, rock pools and larger flowing streams; seeps, springs and smaller creeks; freshwater claypans; birridas; and Lake MacLeod. Environmental factors to which invertebrates appeared to respond were ratio of calcium/ alkalinity, total dissolved solids, turbidity, colour, flow, longitude and nutrients, although some factors were inter-correlated.Additional surveys should find extra species of waterbird and, more particularly, aquatic invertebrate using wetlands of the southern Carnarvon Basin. For many invertebrates, occurrences are too sparse for effective protection of species within a nature reserve system and other mechanisms will be required to ensure their conservation. Comparison of site classifications based on waterbird, aquatic invertebrate and plant data showed patterns among sites identified using one element of the biota did not reflect patterns shown by other elements. This suggests that, until further work has identified an element that reflects the whole wetland community, as many biotic elements as possible should be surveyed.