Somatic hypermutation of immunoglobulin genes occurs in many vertebrates including sharks, frogs, camels, humans and mice. Similarities among species reveal a common mechanism and these include the AGC/T sequence hot spot, preponderance of base substitutions, a bias towards transitions and strand bias. There are some differences among species, however, that may unveil layers of the mechanism. These include a G:C bias in frog and shark IgM but not in nurse shark antigen receptor (NAR), a high frequency of doublets in NAR hypermutation, and the co–occurrence of somatic hypermutation with gene conversion in some species. Here we argue that some of the similarities and differences among species are best explained by error–prone DNA synthesis by the translesion synthesis DNA polymerase zeta (Pol ζ) and, as suggested by others, induction of DNA synthesis by DNA breaks in antigen receptor variable genes. Finally, targeting of the variable genes is probably obtained via transcription–related elements, and it is the targeting phase of somatic hypermutation that is the most likely to reveal molecules unique to adaptive immunity.