Sexual differences in physiology are widely regarded as potential proximate mechanisms that underlie sex differences in mortality, life history and disease risk of vertebrates. However, little is known about the causes of sex-specific variation in physiology. Sexual selection and parental workload are two key components suggested to play a role. Theory predicts that, within males, species with stronger male sexual selection (greater sexual dichromatism and more frequent social polygyny) and higher male parental effort should have lower immune capacity and stronger oxidative imbalance. Within females, weak or no direct effect of male sexual selection on physiology is expected, but species where females invest more in parental care should have lower immune capacity and higher oxidative imbalance. We tested these predictions by phylogenetic comparative analyses conducted separately for the two sexes and based on 11,586 physiological measurements of samples collected in the field from 2,048 individuals of 116 and 106 European species for males and females, respectively. For males, we found that the degree of dichromatism, polygyny and male parental effort correlated negatively with multiple immune indices, and the level of antioxidant glutathione correlated positively with polygyny score. In contrast, female immune and oxidative variables were unrelated or weakly related to both male sexual selection or female parental effort. We conclude that sex roles can drive inter-specific variation in immune function (primarily in male birds), but less so in oxidative physiology. These findings support earlier claims that males pay higher physiological costs of sexual selection than females, but apparently also of caregiving. We discuss how females might avoid such costs.