Most biologists implicitly define an individual organism as "one genome in one body." This definition is based on physiological and genetic criteria, but it is problematic for colonial organisms. We propose a definition based instead on the evolutionary criteria of alignment of fitness, export of fitness by germ-soma specialization, and adaptive functional organization. We consider how these concepts apply to various putative individual organisms. We conclude that complex multicellular organisms and colonies of eusocial insects satisfy these three criteria, but that, in most cases (with at least one notable exception), colonies of modular organisms and genetic chimeras do not. While species do not meet these criteria, they may meet the criteria for a broader concept--that of an evolutionary individual--and sexual reproduction may be a species-level exaptation for enhancing evolvability. We also review the costs and benefits of internal genetic heterogeneity within putative individuals, demonstrating that high relatedness is neither a necessary nor a sufficient condition for individuality, and that, in some cases, genetic variability may have adaptive benefits at the level of the whole.