Background Changes in host phenotype following parasite infection are often considered as host manipulation when they seem advantageous for the parasite. However, putative cases of host manipulation by parasites are rarely tested in field-realistic conditions. Infection-induced phenotypic change cannot be conclusively considered as host manipulation if no evidence shows that this trait is adaptive for the parasite in the wild. Plasmodium sp., the parasites causing malaria in vertebrates, are hypothesized to “manipulate” their host by making their odour more attractive to mosquitoes, their vector and final host. While this is fairly well supported by studies on mice and humans, studies focusing on avian malaria give contradictory results. Methods: In the present study, genotyped birds at different stages (uninfected, acute and chronic) of Plasmodium relictum infection were exposed, in a large outdoor aviary, to their natural vector, the mosquito Culex pipiens . Results: After genotyping the blood meals of more than 650 mosquitoes, we found that mosquitoes did not bite infected birds more than they bit them before infection, nor more than they bit uninfected hosts. Conclusions: Our study highlights the importance of testing ecological behaviours under natural conditions and suggests that different processes might be at play in mammals and birds regarding potential manipulation of attractiveness by malaria parasites.