In an auempt to know the feeding preferences of phlebotomine sandflies that transmitrnvisceral leishmaniasis (VL) in Eth iopia, Phlebotomus mani"i and P. orielllaiis. and thusrnto poilll to the possible reservoir host(s) of the disease in the country, blood meals ofrnthe vectors were identified using the counter current immunoelectrophoresis (CCIE)rntechnique. First, genus/species-specific antisera were raised against the immunoglobulinrnG (lgG) of ten potential hosts of the sand fly vectors: human, sheep, goat, donkey ,rnhorse, camel, cattle, squirrel, mongoose and hyrax. The IgG was fractionated from thernwhole sera of the mammals by the ammonium sulphate precipitation method followedrnby the ion-exchange chromatography technique. The antisera were developed byrnprim ing (subcutaneously) rabbits, in triplicates, with water-in-oil emulsion of equalrnvolumes of IgG and Freund's Complete Adjuvant and then boosting with Freund 'srnIncomplete Adjuvant. The potency of the ant isera was determined by assay ing each lestrnbleed from each rabbit againstlhe homologous IgG. Cross-reactiv ities were checked byrnscreening each test bleed against heterologous IgGs. Human whole serum and IgG wererncross-reacted with anti-cattle antisera and vice versa. All other antisera were specific.rnThe titre of each test bleed was qualitatively determined by visualizing thernconspicuousness of the precipitin line. Hyperimmune antisera were used in blood mealrnanalysis. Eight blood meal samples (5 SergelJlomyia spp. and 3 P. manilll) wererncollected , by sticky and CDC light traps, from Aba-Roba in a total of 42 night catches.rnOne hundred and six blood meal samples (94 P. orielJlalis and 12 P. bergerofl)rncollected, by CDC, from the Middle Awash va lley and stored for about 4 years werernused. Each blood meal sample was tested against cortlmercial anti-dog and anti-rat IgGsrnvirnin addition to the 10 ami-host sera raised in the present study. Of 114 blood mealsrnprocessed, hosts were identified for the 93 (8 1.6%): 2 P. martilli, 79 P. oriellfalis andrnall P. bergeroli. None of Sergelllomyia blood meals were identified. 37.6% of the bloodrnmeals detected were from single host sources: 20 from cattle, 10 camel, 2 squirrel , Irndonkey, I human and I from mongoose. The rest 62.4% was from mixed sources.rnAltogether, 80.6% of the meals identified were from caule origin and 59% from camel,rnsingly or in combination with OIher hosts. Thus. P. oriellfatis appears to be anrnopportunistic feeder with a preference for cattle and camel at least in the Middle Awashrnwhere these hosts occur in large numbers. The CCIE technique is specific and sensi tivernenough not only to detect mixed blood meals from close ly related hosts but halfdigestedrnmeals. The assay is technically simple and inexpensive in terms of reagentsrnrequired. Thus, CCIE is a robust technique for identification of blood meals ofrnhaemalophagous insect vectors in general and smaller nies such as phlebotominernsandfl ies and of biting midges in particular.