Entomological monitoring in four villages situated along an altitude transect in the Hai District of Northeastern Tanzania identified Anopheles arabiensis Patton as the principal vector of malaria and detected seasonal changes in vector behavior. Over a 13-mo sampling period, 10,557 mosquitoes were collected with CDC light traps, pyrethrum spray catches, and pit traps of which 5,969 (56.5%) wereAn. arabiensis, 762 (7.2%) wereAnopheles funestus Giles s.l., 3,578 (33.9%) were culicines, and 248 (2.3%) were nonvector anophelines. Vector densities declined rapidly with increasing altitude, demonstrating a 50% decrease in annual human biting rate for every 86-m rise in altitude. Light traps were found to be more efficient than spray catches for the collection of An. arabiensis. This observation was attributed to increased exophily of this species, most notably in the wet season, and is supported by seasonal changes in the human blood index and fed/gravid ratio. These results indicate that spray catches may underestimate the abundance of exophilic vectors such as An. arabiensis and that entomological monitoring may require more than one collection method, especially at low vector densities. The annual entomological inoculation rate (EIR) decreased sharply with increasing altitude, with large variation around the estimate at low vector densities. Increased transmission because of unpredictable short rains at low altitudes and spatial clustering of infective mosquitoes may contribute to elevated EIR estimates.