If DNA replication is a major cause of mutation, then those life-history characters, which are expected to affect the number of male germline cell divisions, should also affect the male to female mutation bias (α_m). We tested this hypothesis by comparing several clades of bird species, which show variation both in suitable life-history characters (generation time as measured by age at first breeding and sexual selection as measured by frequency of extrapair paternity) and in α_m, which was estimated by comparing Z-linked and W-linked substitution rates in gametologous introns. α_m differences between clades were found to positively covary with both generation time and sexual selection, as expected if DNA replication causes mutation. The effects of extrapair paternity frequency on α_m suggests that increased levels of sexual selection cause higher mutation rates, which offers an interesting solution to the paradox of the loss of genetic variance associated with strong directional sexual selection. We also used relative rate tests to examine whether the observed differences in α_m between clades were due to differences in W-linked or Z-linked substitution rates. In one case, a significant difference in α_m between two clades was shown to be due to W-linked rates and not Z-linked rates, a result that suggests that mutation rates are not determined by replication alone.