Abstract
This paper begins by asking to what extent numbers of species are an adequate measure of biological diversity, either locally or globally; both for evolutionary understanding and for practical applications, biodiversity may often be better quantified at lower or higher levels, from genes to ecosystems. The subsequent discussion, however, focuses on species, and discusses questions that arise in estimating how many species there have ever been, how many there currently are in various taxonomic groups, and how we may quantify the differing degrees of `independent evolutionary history' or `taxonomic distinctiveness' in different species or groups. I conclude with opinions about how the practical task of identifying and recording species diversity might be better managed.