The identification of freshwater macroinvertebrates to family level is becoming increasingly popular for surveys, predictive models and pollution indices in Australia because it is quicker and cheaper than genus or species level identification, and it requires less specialised knowledge. Family richness has been used as a predictor of species richness for other taxonomic groups such as vertebrates, ants and plants and we were interested in seeing whether this might be a useful method for freshwater macroinvertebrates. Taxon lists from one Papua New Guinean and 34 Australisn datasets from lentic and lotic waters were used to regress the number of families against the total number of taxa (species where possible). Also, the ability of the number of species and morphospecies within some orders (Coleoptera, Diptera, Ephemeroptera and Trichoptera) to predict overall species richness was investigated. The number of all families explained 91% of the variation in species richness. The number of species within each of the ordersexplained between 60 and 85% of the variation in overall species richness. We conclude that it would be posiible to predict species richness in this way, particularly if sampling techniques and sampling effort were standardised. The terms 'Species richness' and 'Biodiversity' are often used synonymously although the former is only a subset of the latter. Some of the limitations and dangers of assessing species richness instead of biodiversity are discussed.