Biogeographers often investigate patterns of biodiversity at continental and global scales, using existing data georeferenced to a lattice of cells of latitude and longitude. Problems can arise with this approach when the available biological data are insufficient to adequately sample each cell and the cells are environmentally heterogeneous. An alternative, though less-often employed, approach is to use bioregions (defined as areas with distinctive biophysical environmental characteristics) as the basic sampling unit and to statistically control for unequal areas of regions. Here we applied this latter approach with the Interim Biogeographical Regionalisation of Australia (IBRA) to analyse continental patterns of songbird species richness in relation to mean annual precipitation, mean annual temperature, and mean wet season temperature, which are all predicted to substantially change given anthropogenic climate change. We used the Birds Australia database that has a large sample (> 1,560,000) of distribution records covering Australia. For each of the 85 IBRAs, we determined the total number of songbird species and standardized these richness values accounting for the species-area effect by including the log of bioregion area as a covariate in the statistical models. Our analysis of standardized bioregion songbirds richness showed that the best supported model, based on information theory statistics included an interaction of mean annual temperature and precipitation (48.6% deviance explained). The fitted model showed declining richness with increasing temperature and declining precipitation, signalling that future climates may result in regional declines in songbird abundance. We suggest our simple empirical-statistical approach, using bioregions as the spatial unit, has promise for continental and global impact assessment of diversity changes and for conservation planning.