Background: The majority of P. vivax and P. falciparum infections in low-endemic settings are asymptomatic. The relative contribution to the infectious reservoir of these infections, often of low-parasite-density, compared to clinical malaria cases, is currently unknown but important for malaria elimination strategies. Methods: We assessed infectivity of passively-recruited symptomatic malaria patients (n=41) and community-recruited asymptomatic individuals with microscopy- (n=41) and PCR-detected infections (n=82) using membrane feeding assays with Anopheles arabiensis mosquitoes in Adama, Ethiopia. Malaria incidence and prevalence data was used to estimate the contributions of these populations to the infectious reservoir. Results: Overall, 34.9% (29/83) of P. vivax and 15.1% (8/53) P. falciparum infected individuals infected ≥1 mosquitoes. Mosquito infection rates were strongly correlated with asexual parasite density for P. vivax (ρ = 0.63; P < .001) but not for P. falciparum (ρ = 0.06; P = .770). P. vivax symptomatic infections were more infectious to mosquitoes (infecting 46.5% of mosquitoes, 307/660) compared to asymptomatic microscopy-detected (infecting 12.0% of mosquitoes, 80/667; P = .005) and PCR-detected infections (infecting 0.8% of mosquitoes, 6/744; P < .001). Adjusting for population prevalence, symptomatic, asymptomatic microscopy- and PCR-detected infections were responsible for 8.0%, 76.2% and 15.8% of the infectious reservoir for P. vivax, respectively. For P. falciparum, mosquito infections were sparser and also predominantly from asymptomatic infections. Conclusions: In this low-endemic setting aiming for malaria elimination, asymptomatic infections are highly prevalent and responsible for the majority of onward mosquito infections. The early identification and treatment of asymptomatic infections might thus accelerate elimination efforts.