National estimates of HIV prevalence of the whole adult population are based onrnANC based sentinel surveillance system. Therefore with the objective of determiningrnthe representativeness of the ANC attendees in relation to pregnant women notrnattending ANC and general adult population, a cross sectional comparative studyrnwas conducted. It employed a pre-tested standardized questionnaire to collectrninformation on risk factors related to HIV/AIDS from a sample of 1426 individuals.rnData was collected in Addis Ababa on April 2000 at the four health centers where thernsentinel surveillance is ongoing and at the respective catchmentâ€™s areas of the healthrncenters. Forty seven percent of the attendees belonged to the age group 15 - 24,rn72% were unemployed, 33% had a monthly income of 500 Birr or more, 36% werernnew comers to the community and 89% were married. Although 61% werernmultipara, 50% had no children. Over 52% married at age below 19 years, 83.6%rnnever used condom.rnThere were no significant difference in risk profile between ANC attendees and thernnon-attendees pregnant women while there were significant differences among ANCrnattendees and the general adult population of females. ANC attendees were youngerrnwith low education; unemployed (72%), new comers to the area, and marriedrncompared to the general population of females. Majority of ANC attendees neverrnused condom with regular and non-regular partners (92.7% and 77.6% respectively)rnand reported STI was 16.4%.rnMales were found to have significant association with risk factors i.e. majority of themrnwere single (71.4%), had sexual intercourse with high risk partner (55%) with onlyrnviiirn33% of every time use of condom with such partners. The related risk factors werernobserved to distribute significantly between ANC attendees and the GP of malesrnshowing both to have a high-risk profile. It was then concluded that ANC attendeesrncan confidently be taken as representative of the general pregnant population andrnthe GP of males, differ significantly with the general population of female who werernobserved to exhibit a low risk profile. In addition, close to 30% sexually inactivernrespondents in the GP, 52% never pregnant and 55% childless women werernobserved, an addition of another sentinel population as an adjunct to the existing onernin order to capture these population is recommended.