Baseline interleukin-7 (IL-7) level has been proposed as a predictor of immunologic response to antiviral therapy, and the use of exogenous IL-7 has been suggested in the treatment of HIV-infected patients. Therefore, we investigated the relationships between IL-7 serum levels and immunologic outcome of highly active antiretroviral therapy (HAART) in 24 children perinatally infected with HIV-1. IL-7 serum levels were evaluated by enzyme-linked immunosorbent assay before switching antiretroviral treatment from double therapy to HAART and then again after 12 weeks of HAART. Differences were analyzed between children with immunologic failure and those without. At baseline, IL-7 levels were significantly higher in HIV-1-infected children than in 24 healthy age-matched uninfected children (16.6 +/- 8.7 vs. 9.5 +/- 2.4 pg/mL, respectively; P < 0.001). IL-7 levels inversely correlated with CD4 T-lymphocyte percentage both at baseline (r = -0.71, P < 0.001) and after 12 weeks of HAART (r = -0.49, P = 0.01). We observed no correlation between IL-7 levels and viral load (r = 0.29, P = 0.17 at baseline; r = 0.30, P = 0.15 after 12 weeks). Baseline IL-7 levels were similar in HIV-1-infected children with or without subsequent immunologic failure (15.1 +/- 8.8 vs. 17.5 +/- 8.7 pg/mL, respectively; P = 0.75). After 12 weeks of HAART, IL-7 levels were 21.6 +/- 13.5 pg/mL in children with immunologic failure (P = 0.08 when compared with baseline levels) and 8.2 +/- 3.8 pg/mL in children without immunologic failure (P = 0.002 when compared with baseline levels). IL-7 levels were significantly higher (P = 0.003) in children with immunologic failure than in those without. Findings indicate that baseline IL-7 serum levels do not predict immunologic outcome of HAART and that immunologic failure occurs even in the presence of high IL-7 serum levels, thus suggesting no benefit from therapy with exogenous IL-7.