Abstract
The evolution of dominance has been subject to intensive debate since Fisher ®rst argued that modi®ers would be selected for if they made wild-type alleles more dominant over mutant alleles. An alternative explanation, put forward by Wright, is that the commonly observed dominance of wild-type alleles is simply a physiological consequence of metabolic pathways. Wright's explanation has gained support over the years, largely ending the debate over the general recessivity of deleterious mutations. Nevertheless there is reason to believe that dominance relationships have been moulded by natural selection to some extent. First, the metabolic pathways are themselves products of evolutionary processes that may have led them to be more stable to perturbations, including mutations. Secondly, theoretical models and empirical experiments suggest that substantial selection for dominance mod-i®ers exists during the spread of adaptive alleles or when a polymorphism is maintained either by overdominant selection or by migration-selection balance.