Background: Bacterial vaginosis is one of the most common genital tract infectionsrnamong reproductive age group. It is associated with different gynecologic and poor obstetricrnout come. The prevalence of bacterial vaginosis varies from country to country even in thernsame country it varies at population of interest. Different social and sexual factors canrncontribute for development of bacterial vaginosis. Bacterial vaginosis can be easilyrndiagnosed by combination of two Amsel’s criteria.rnObjectives: This study was undertaken to determine the prevalence of bacterial vaginosisrnand to evaluate the accuracy of Amsel’s criteria individually or in combination of two forrnthe clinical diagnosis of bacterial vaginosis among pregnant women attending Antenatal carernin Tikur Anbessa University Hospital, Addis Ababa, Ethiopia. An attempt has also beenrnmade to identify the possible risk factors associated with bacterial vaginosis.rnMethods: In this observational study during the period from November 2011 to April 2012,rn252 pregnant women were screened for bacterial vaginosis in Tikur Anbessa UniversityrnHospital. Vaginal swabs were collected for pH determination, saline wet mount microscopicrnexamination to detect clue cells, KOH preparation for whiff test and Gram-stain evaluationrnof vaginal flora for diagnosis of bacterial vaginosis by Nugent scoring system. Accuracy ofrnclinical diagnosis using individual and two of Amsel’s criteria was evaluated.rnResults: The prevalence of bacterial vaginosis was 18.3% by Amsel’s two of three criteriarnand 19.4% using Gram stain Nugent scoring system. The prevalence of bacterial vaginosisrnwas 31.6% and 15.9% among symptomatic and asymptomatic pregnant women respectively.rnA high percentage of bacterial vaginosis positive pregnant women were asymptomaticrn(63.3%) whereas the remaining 36.7% bacterial vaginosis positive pregnant women wererncomplaining abnormal vaginal discharge with or without unpleasant smell. Comparing withrnGram stain Nugent scoring methods the clinical diagnosis by Amsel’s criteria had sensitivityrnof 85.7%, specificity of 98%, and positive predictive value of 91.3% and negative predictivernvalue of 96.6%. The most sensitive and specific individual criterion was clue cells and withrnhighest positive and negative predictive value. Amsel’s criteria with the lowest sensitivityrnand specificity were whiff test and vaginal pH respectively. Combination of clue cells withrnvaginal pH test were the highest in sensitive while whiff test with clue cells were the highestrnin specificity than the other combined two Amsel’s criteria. After adjusted for other factorsrnmultiple life time sexual partner (OR: 8.6; 95% CI: 2.5, 29) and previous history ofrnspontaneous abortion (OR: 5.9; 95% CI: 1.5, 23) had remained significantly associated withrnprevalence of bacterial vaginosis. They were the most likely to be a risk factors for bacterialrnvaginosis infection.rnConclusion and Recommendation: The prevalence of bacterial vaginosis among pregnantrnwomen was higher in asymptomatic pregnant women and associated with the factorsrnmultiple lifetime sexual partner and previous spontaneous abortion. Amsel’s criteriarndiagnosis of bacterial vaginosis can be simplified using a combination of the two criteria,rnvaginal pH and clue cells, in settings where time or Gram staining is not available. So usingrnthese simplified diagnostic criteria is better to screen pregnant women as a routine activityrnduring antenatal care.