Frequency distributions of body size have been reported in the literature for a range of animal higher taxa. However, the reported shapes of these distributions may be biased by species missing from them, specifically by currently undiscovered species. There is a body of evidence that the small-bodied species in a taxon are described later, on average, than the large-bodied. From this, we predict that the means of animal body size frequency distributions should decrease through time, and the skewnesses of the distributions increase. These predictions are shown to be true for body size distributions for five different higher taxa of animals. Thus, any reported body size distribution for a taxon is likely to be systematically biased, even if it includes all the species which are known, if many species are likely to remain undiscovered.