In the ongoing debate concerning the nature of human racial categories, there is a trend to reject the biological reality of race in favour of the view that races are social constructs. At work here is the assumption that biological reality and social constructivism are incompatible. I oppose the trend and the assumption by arguing that cladism, in conjunction with current work in human evolution, provides a new way to define race biologically. Defining race in this way makes sense when (...) compared to the developments in other areas of systematic biology, where shared history has largely replaced morphological similarity as the foundation of a natural biological classification. Surprisingly, it turns out that cladistic races and social constructivism are compatible. I discuss a number of lessons about the way human biological races have been conceptualized. (shrink)