Three hun dred fi ft y o ne (J 95 loca l zebu and 156 Ho lste in x Loca l Zebu) la ctating cows o f rnsma ll holde r p ri va te fa rm s in Ba hir Dar milk shed were examined fro m Septe mber, 2003 to rnMa rch, 2004 ( 0 de te rm ine mastiti s prevale nce, isolate pa thogens in vo lved, eva lua te the rnant ibiotic susceptib ili ty pro fi les a nd [ 0 eval ua te soma ti c cell count in identify in g rnintrama mmary in fec ti o ns. Clinical preva lence was deteml in ed th rough examinati o n of rnabno mla lities of m ilk , udde r or cow. Cal ifornia mastitis tes t (C MT) and c ulture were used for rnsubclinical mastiti s de te rm ination. Aga r disc di ffusion was used fo r antibiotic suscept ibi lity rntest. Somati c cell count was condu cted foll owing standard procedures desc ri bed In rnInte rn ati o nal Dairy Federdt io n for e num eratio n of ce ll s wi th direc t Mi croscopic me thod. rnClinical prevalence at cow level was 3.9% in crossbreds and no nc in local zebu breeds. rnSubclinical mastit is a t cow level based o n C MT was high (34.4%) in crossbreds compared to rnindigenous zebu (17.9%) (p < 0.05). Quarter subclinical prevale nce based on eMT was rn17 .90% and 4.95% for c rossbreds and local zebu. respectively. Amo ng potential risk fac tors rnconsidered, stage of lactation, parity and breed were found 10 afTecI the occurrence of mastitis rnsignificantly (p < 0.05). The pathogens isola ted from ma stitic milk were coagulase negalive rnstaphylococci (eNS), s. aureus. Srr. agalacliae. Sir dysgalactiae. Sr r. uberis. Micrococcus rnspecies, C. bovls. A. pyogefls, B. cereus, and S. illtermedius. Among these, the most frequent rnisolates we re eNS (49 .6% of the 10la l isolates). S. alirellS (17.9%), Srr. agalacliae (8.2%) and rnSIr dysgalacliae (6:7%). rnSeven antibiotics including sul fi soxazole. tetracycline, erythromycin. oxacil lin, rnchloramphenicol, cJindamycin, and streptomycin were tested o n 81 isola tes. Except fo r rnstreptomycin, all isolates ere sensitive to all antibiotics. All isolates were most se nsi tive 10 rnsulfisoxazo le. Staphylococcw; allreus was susceptible to all drugs except streptomycin. rnSllIphylococcus Ollreus was 91.7% susceptib le to oxacillin, however, CNS were less rnsusceptible (68.2%) ill vilro. rnTo evaluate somatic cell count (SeC) in ide nt ify ing imramamary infect ion fo r crossbreds. 10 rncutoff poims between 100000 cellsl ml and 300000 celJSJml were evaluated for the ir rnsensllivity and speci-ficity in comparison with cu ltural results. Sim ilarly, cutoff points between rn80000 cell ml and "250000 cell ml were take n fo r loca l zebu breeds. The sensitivity for rnVI rncrossbreds range from 79.16 % (al 300000 ce ll slml cutoff point) 10 95.80 % (al 100000 cells! rnml) and similarly specificity from 80.28% to 45.52%. Fo r local zebu breeds in the same order, rnsensitivity range from 58.82% to 88.23% and specificity from 45.82% to 85.15%. To rnestablish threshold level this study was a first anempl in Ethiopia and to usc SCC as a rndiagnostic tool on a national sqle further study need to be conducted with improved cultural rntechnique and automatic cell counters. Cross breed cows had high see compa red to local rnzebu cows. The sec was significantly (p