There is little direct evidence of the fitness effects of changes in malaria gametocyte sex ratio. Gametocyte sex ratios in haemospororin parasites (phylum Apicomplexa) are usually female skewed. However, in some cases and especially in Haemoproteus parasites, less female–biased and even male–biased sex ratios are encountered. The ‘fertility insurance hypothesi’ tries to explain these biases as an evolutionary strategy to facilitate gamete encounter. Thus, the hypothesis predicts that, if there is a reduction in gametocyte density (intensity of infection) or other factors preventing gametes from meeting, a change to a higher proportion of male gametocytes may be favoured. By contrast, a change in sex ratio may be caused by other non–adaptive mechanisms, for example differential survival of the gametocytes of each sex. We study within–host changes in Haemoproteus majoris sex ratios following an experimental reduction in the density of the parasites in the blood in a breeding population of blue tits (Parus caeruleus). Medication with the antimalarial drug primaquine induced a significant reduction in Haemoproteus gametocyte infection intensity in two different breeding seasons and under two different doses of medication. Sex ratios became male skewed following the experimental treatment in agreement with the predictions of the‘fertility insuranc’ hypothesis. Also in support of the hypothesis, a significant change towards male–biased sex ratios emerged for non–medicated birds in one year, probably owing to the natural immune reduction of the density of the parasites in the blood. The alternative possibility that changes are caused by different lifespans of gametocytes is not supported by changes in sex ratios in control hosts, where new production and release of gametocytes occur.