Few studies have evaluated the extent of human immunodeficiency virus (HIV) in the adolescent population. However, there has been growing concern that sexual and drug experimentation common in this age group may increase their risks of transmitting the virus. Between October 1, 1987, and January 31, 1989, a blinded, unlinked HIV seroprevalence study was conducted among all adolescents aged 13 through 19 receiving ambulatory care at Children's National Medical Center and having blood drawn for other routine medical indications. Overall, seroprevalence in this group of patients was 0.37% (3.7/1000), with the highest prevalence in females (4.7/1000) and patients 18 through 19 years of age (5.6/1000). Of adolescents considered at high risk who were offered and accepted voluntary HIV testing during the same time period, 4.1% (41/1000) were positive. Inasmuch as this represents only 38% of all of the positive tests obtained in the blinded testing phase of the study, it may indicate that a substantial proportion of HIV-positive adolescent patients may be missed by using standard criteria and methods of identifying risk and/or that those most at risk may be reluctant to be tested for HIV infection. The results suggest that HIV infection is present in this population of urban adolescents and that the seroprevalence rate is higher than in other nonselect groups. Moreover, using traditional risk factors as screening criteria may not identify the majority of those infected. Trends need to be followed and further studies conducted in an attempt to define which adolescents are at highest risk for HIV infection.