Arboviruses constitute a major public health problem, in particular mosquito-borne arboviruses that continue to emerge and re-emerge. Arbovirus infection of mammals is enhanced by the presence of a mosquito bite at the inoculation site, in comparison to virus experimentally administered by needle inoculation in the absence of a bite. Inflammatory responses to bites appear to be a key factor in this enhancement. However, the experimental inoculation of mosquito saliva with virus inoculum by needle, in the absence of bite trauma, also has the ability to enhance viral infections. In this study, we have studied the mechanistic basis for these observations. We have studied whether saliva from different mosquito species can enhance virus infection. Interestingly, while saliva from Aedes genus enhanced virus infection, An. gambiae saliva did not. Instead, An. gambiae saliva actively inhibited infection compared to inoculation with virus alone. This could partly explain why An. gambiae mosquitos are unsuitable vectors for transmitting most arboviruses. By comparing the effects that saliva of these different mosquitoes have at the bite site we have further specified which inflammatory responses at the inoculation site modulate arbovirus infection in the skin. Using an in vivo mouse model we have shown that An. gambiae causes significantly less oedema but a higher up-regulation of key inflammatory genes in the skin than A. aegypti. As such, we are providing important insights into how mosquito saliva modulates infection. A better understanding of this will aid the development of anti-viral treatments by targeting factors within the mosquito bite that are common to many distinct infections.