Immunosuppressive effects of testosterone lead to a prediction of male
-biased parasitism. To test this prediction, prevalences of blood para
sites were compared between male and female birds using statistically
correct vote counts of data From 33 studies. We found no overall diffe
rence in prevalence between males and females, in either breeding or n
on-breeding birds. However, infections by Haemoproteus (the most commo
n genus of blood parasite found) were significantly more common among
breeding females than breeding males. Restricting the analysis to bree
ding birds of polygynous species, females again were more likely than
males to be infected by blood parasites: this result held for an intra
-family comparison that controlled for phylogenetic effects. In compar
ison, measures of sexual size dimorphism did not relate to sex biases
in parasitism as predicted, after controlling for phylogeny using inde
pendent comparisons. Because testosterone is often implicated in suppr
essing the immune system, female biases in parasitism are unexpected.
Female biases in parasitism by blood parasites could result from diffe
rential exposure of the sexes to vectors, or from oestrogen-based effe
cts on immunity.