In much of sub-Saharan Africa, the mosquito Anopheles gambiae is the main vector of the major human malaria parasite, Plasmodium falciparum. Convenient laboratory studies have identified mosquito genes that affect positively or negatively the developmental cycle of the model rodent parasite, P. berghei. Here, we use transcription profiling and reverse genetics to explore whether five disparate mosquito gene regulators of P. berghei development are also pertinent to A. gambiae/P. falciparum interactions in semi-natural conditions, using field isolates of this parasite and geographically related mosquitoes. We detected broadly similar albeit not identical transcriptional responses of these genes to the two parasite species. Gene silencing established that two genes affect similarly both parasites: infections are hindered by the intracellular local activator of actin cytoskeleton dynamics, WASP, but promoted by the hemolymph lipid transporter, ApoII/I. Since P. berghei is not a natural parasite of A. gambiae, these data suggest that the effects of these genes have not been drastically altered by constant interaction and co-evolution of A. gambiae and P. falciparum; this conclusion allowed us to investigate further the mode of action of these two genes in the laboratory model system using a suite of genetic tools and infection assays. We showed that both genes act at the level of midgut invasion during the parasite's developmental transition from ookinete to oocyst. ApoII/I also affects the early stages of oocyst development. These are the first mosquito genes whose significant effects on P. falciparum field isolates have been established by direct experimentation. Importantly, they validate for semi-field human malaria transmission the concept of parasite antagonists and agonists.