HIV infection is routinely diagnosed by various immunoassays that detect the presence of anti-rnHIV antibodies. But these methods do not distinguish recent from established infections. Thernpreviously devised method called Ig G-capture BED-EIA (Subtype B, E, D, Enzyme ImmunornAssay) is used to distinguish new from established infection based on standardized opticalrndensity values, which is a measure of antibody titer in the serum using cross-sectional specimen.rnThe objective of this study was to identify recent HIV-1 infection using STARHS assay andrnestimate HIV-1 incidence among pregnant women attending antenatal care clinics in AddisrnAbaba and to compare the performance of 'in -house' incidence assay kit with commercial assayrnusing 1078 cross-sectional samples collected from pregnant women attending Antenatal CarernClinics (ANC) based on Standardized Optical Density (OD-n) values. Larger percentage of recentrninfection was observed among women attending the inner city health centers as compared to thernouter city health centers and among women aged 1 5 - 1 9 years. A decline in HIV incidence wasrnobserved in recent years among women of young age groups attending the inner city healthrncenters. Significant correlation was observed between the two assays. There were 84 commonrnsamples which had an OD-n value of less than 0.8 and 954 samples with standardized opticalrndensity of greater or equal to 0.8 when tested by both commercial and house -made kits. But fortyrnsamples (3.7%) had discrepant results. An agreement was also observed between the OD-n valuesrnof the two assays, with a correlation coefficient (r) of 0.76.