This retrospective study delineates outcome status at age three by AIDS defining diagnosis for children with perinatally acquired HIV infection, and determines whether prenatal drug exposure and preterm birth are related to outcome status. The study sample consists of HIV-infected children who were attended by the Special Immunology Service and who had at least one developmental evaluation (Bayley or McCarthy Scales)during the study period(February 1991-May 1995). We selected a limited age range to determine neurocognitive outcome,(3 years +/- 6 months), distinguished outcome by AIDS defining illness (LIP and non-LIP, and incorporated mortality into a definition of extremely poor outcome status. We hypothesized that prenatal drug exposure and low birthweight would negatively impact outcome. Seventy-four children met the CDC criteria for diagnosis of HIV infection; 48 had AIDS (1987 AIDS surveillance case definition); 16 had LIP AIDS. All had receeived antiretroviral treatment. Twelve children were not available for testing within the age limit selected for study and were excluded. Twelve additional children died before they could be tested at age 3. The remaining 50 children all received neurocognitive testing at age 3 (+ 6 months). The majority were functioning within broadly normal limits on neurocognitive testing at age three (DQ >70). However, 53% of the children with a diagnosis of AIDS had neurocognitive test scores <70. Children who had LIP AIDS were less likely to score in that range than those with non-LIP AIDS, although more likely than HIV+ children who did not have AIDS (p=.009). Extremely poor outcome, as defined by the lower limit of the Bayley or McCarthy test (<50) or death by age three, occurred exclusively among those with a diagnosis of non-LIP AIDS. Lower birthweight was associated with extremely poor outcome status, although prenatal drug exposure was not. Outcome studies of HIV-infected children are a reflection of the survival of those less affected with the disease.