The induction of cellular immunity, in conjunction with antibodies, may be essential for vaccines to protect against blood-stage infection with the human malaria parasite Plasmodium falciparum. We have shown that prime-boost delivery of P. falciparum blood-stage antigens by chimpanzee adenovirus 63 (ChAd63) followed by the attenuated orthopoxvirus MVA is safe and immunogenic in healthy adults. Here, we report on vaccine efficacy against controlled human malaria infection delivered by mosquito bites. The blood-stage malaria vaccines were administered alone, or together (MSP1+AMA1), or with a pre-erythrocytic malaria vaccine candidate (MSP1+ME-TRAP). In this first human use of coadministered ChAd63-MVA regimes, we demonstrate immune interference whereby responses against merozoite surface protein 1 (MSP1) are dominant over apical membrane antigen 1 (AMA1) and ME-TRAP. We also show that induction of strong cellular immunity against MSP1 and AMA1 is safe, but does not impact on parasite growth rates in the blood. In a subset of vaccinated volunteers, a delay in time to diagnosis was observed and sterilizing protection was observed in one volunteer coimmunized with MSP1+AMA1-results consistent with vaccine-induced pre-erythrocytic, rather than blood-stage, immunity. These data call into question the utility of T cell-inducing blood-stage malaria vaccines and suggest that the focus should remain on high-titer antibody induction against susceptible antigen targets.