The ecology and behaviour of Anopheles mosquitoes was studied in four selected sites inrnthe Metehara area, Upper Awash Valley, Eastern Ethiopia. The sites represented two insecticidernunsprayed villages (Metehara and Gelcha) and two sprayed villages (the Sugar Estate and Buse).rnInformation on the prevalence of malaria cases was also gathered from the Metehara SugarrnEstate Hospital and the East Shoa Malaria Control Sector (Nazareth). The results showed that arntotal of 24,799 microscopically diagnosed malaria cases out of 68,000 blood samples werernregistered between July 1999 and September 2000 in the two health service renderingrnorganizations. Plasmodium falciparum and P. vivax were found to be responsible for the diseasernand occur nearly in equal proportions (52.7% P. falciparum and 47.3% P. vivax ).rnLarval collections from the different breeding habitats revealed the presence of fourrnspecies of which Anopheles arabiensis was the predominant followed by. An. pharoensis.rnSimilarly, a total of 3639 adult anophelines representing at least eight species were caught usingrndifferent methods from July 1999 to September 2000. An. arabiensis was the predominantrnspecies, forming 95% of all collections followed by An. pharoensis (3.8%).rnThe density of An. arabiensis resting indoors in the sprayed sites was much lower thanrnthe density in the unsprayed sites. Consequently, 18%, 21.3%, 66.1% and 69.4% of An.rnarabiensis exhibited exophily in Gelcha, Metehara, Buse and the Sugar Estate, respectively.rnThe indoor to outdoor biting ratio of An. arabiensis varied between villages: 0.53 in thernSugar Estate, 1.97 in Buse, 0.75 in Gelcha and 2.12 in Metehara town, showing that the speciesrnwas more endophagic in Buse and Metehara and more exophagic in Gelcha and the Sugar Estate,rn*rn:rnXIrnThe highest man biting rate of 39.5, was recorded in Metehara town and the lowest 8.4 in Buse.rnAn. pharoensis was mostly an outdoor feeder.rnOf 864 An. arabiensis and 63 An. pharoensis dissected from human bait collections, thernaverage parous rate was 45.1% and 30.2%, respectively, showing An. arabiensis to be longerlivedrnthan An. pharoensis. The sporozoite rate of An. arabiensis was 0.77% in parous and 0.21%rnin nulliparous population, the overall being 0.46%. Similarly, An. pharoensis had sporozoite raternof 5.3% in parous and 2.3% in nulliparous population, the overall being 3.2%. The biting rhythmrnof An. arabienisis exhibited two to three peaks of activity before and after midnight . The highestrnbiting density occurred after midnight indoors while variation was observed outdoors.rnThe average daily entomological inoculation rate (EIR) of An. arabiensis was 0.05 whilernthat of An. pharoensis was 0.01.rnThe Human blood index (HBI) of An. arabiensis revealed variation betweenrnsites/villages being 1 in Metehara, 0.47 in Gelcha, 0.85 in Buse and 0.93 in the Sugar Estate, thernoverall being 0.65. The HBI varied also between dwelling conditions of mosquito sampling,rnbeing highest in human dwellings (0.78) and lowest in animal shelters (0.13) showing thernopportunistic feeding behaviour of An. arabienisis.rnClearly, An. arabiensis is the most important vector and An. pharoensis a secondaiyrnvector of malaria in the Metehara area.rnInsecticide susceptibility studies in Metehara showed that 30% arid 25% of An.rnrnarabiensis was resistant to DDT and pennethrin , respectively. The level of DDT and permethrinrnresistance in An. arabiensis does not seem to be epidemiologically dangerous, but requiresrnfrequent monitoring. An. arabiensis was found to be highly susceptible to propoxur (carbamaterninsecticide) .