The widespread use of indoor-based malaria vector control interventions has been shown to alter the behaviour of vectors in Africa. There is an increasing concern that such changes could sustain residual transmission. This study was conducted to assess vector species composition, feeding behaviour and their contribution to indoor and outdoor malaria transmission in western Kenya.Anopheles mosquito collections were carried out from September 2015 to April 2016 in Ahero and Iguhu sites, western Kenya using CDC light traps (indoor and outdoor), pyrethrum spray catches (PSCs) (indoor) and pit shelters (outdoor). Species within Anopheles gambiae s.l. and Anopheles funestus s.l. were identified using polymerase chain reaction (PCR). Enzyme-linked immunosorbent assay (ELISA) was used to determine mosquito blood meal sources and sporozoite infections.A total of 10,864 female Anopheles mosquitoes comprising An. gambiae s.l. (71.4%), An. funestus s.l. (12.3%), Anopheles coustani (9.2%) and Anopheles pharoensis (7.1%) were collected. The majority (61.8%) of the anopheline mosquitoes were collected outdoors. PCR result (n = 581) revealed that 98.9% An. arabiensis and 1.1% An. gambiae s.s. constituted An. gambiae s.l. in Ahero while this was 87% An. gambiae s.s. and 13% An. arabiensis in Iguhu. Of the 108 An. funestus s.l. analysed by PCR, 98.1% belonged to An. funestus s.s. and 1.9% to Anopheles leesoni. The human blood index (HBI) and bovine blood index (BBI) of An. arabiensis was 2.5 and 73.1%, respectively. Anopheles gambiae s.s. had HBI and BBI of 50 and 28%, respectively. The HBI and BBI of An. funestus was 60 and 22.3%, respectively. Forage ratio estimate revealed that An. arabiensis preferred to feed on cattle, An. gambiae s.s. showed preference for both human and cattle, while An. funestus preferred human over other hosts. In Ahero, the sporozoite rates for An. arabiensis and An. funestus were 0.16 and 1.8%, respectively, whereas in Iguhu, the sporozoite rates for An. gambiae s.s. and An. funestus were 2.3 and 2.4%, respectively. In Ahero, the estimated indoor and outdoor entomological inoculation rate (EIR) was 108.6 infective bites/person/year (79.0 from An. funestus and 29.6 from An. arabiensis) and 43.5 infective bites/person/year (27.9 from An. arabiensis and 15.6 from An. funestus), respectively. In Iguhu, the estimated indoor and outdoor EIR was 24.5 infective bites/person/year (18.8 from An. gambiae s.s. and 5.7 from An. funestus) and 5.5 infective bites/person/year (all from An. gambiae s.s.), respectively.Anopheles gambiae s.s. showed an increasing tendency to feed on cattle. Anopheles arabiensis was highly zoophagic, whereas An. funestus showed anthropophagic behaviour. While the majority of malaria transmission occurred indoor, the magnitude of outdoor transmission was considerably high. Additional control tools that complement the existing interventions are required to control residual transmission.