Species-range-size distributions have received remarkably little attention in contrast to species-abundance distributions. However, recognition of the importance of regional scale phenomena for local assemblage structure, and the emergence of 'macroecology', have begun to change this situation. A growing number of studies suggests that these distributions are, in general, approximately lognormal, although interpretation is complicated by a variety of factors. Assuming the distribution pattern to be real, it can be viewed in terms of evolutionary and ecological determinants of species occurrences, although their relative significance remains unclear. The form of the distribution has a variety of important consequences, particularly for inventories of faunas and floras and for conservation.