This paper emphasizes the crucial role of variation, at several different levels, for a detailed historical understanding of the development of the biomedical sciences. Going beyond valuable recent studies that focus on model organisms, experimental systems and instruments, we argue that all of these categories can be accommodated within our approach, which pays special attention to organismal and cultural variation. Our empirical examples are drawn in particular from recent historical studies of nineteenth- and early twentieth-century genetics and physiology. Based on (...) the quasi-paradoxical conclusion that biological and cultural variation both constrains and enables innovation in the biomedical sciences, we argue that more attention should be paid to variation as an analytical category in the historiography of the life sciences. (shrink)