Discrepancies between malaria inoculation rates measured entomologically and parasitologically may be explained, at least in part, if infants and children receive less mosquito bites per night than do adults. We found that this problem could be studied by choosing women and children of different ABO blood groups. In preliminary laboratory studies it was found that the blood group of a mosquito's blood meal could be determined in parous and nulliparous mosquitoes for at least 24 hours, and, nullipares up to 34 hours, after feeding. An antiserum against the O group was necessary to distinguish non A or B red cells from those of animal origin. Cross reactions did occur, presumably as a result of the digestion by mosquitoes of the red cell surfaces, but in every case the strongest and earliest developing agglutination was that of the host. Field studies were made using women and children sleeping under mosquito nets, the holes in which made the nets a trapping device. The women, on average, received over seven times more bites per night than did the children. The migration of blood-fed mosquitoes from one net to another was negligible.