BACKGROUND: Data on long-term toxicity of antiretroviral therapy (ART) in HIV-infected children are sparse. PENPACT-1 was an open-label trial in which HIV-infected children were assigned randomly to receive protease inhibitor (PI)- or nonnucleoside reverse-transcriptase inhibitor (NNRTI)-based ART. METHODS: We examined changes in clinical, immunologic, and inflammatory markers from baseline to year 4 in the subset of children in the PENPACT-1 study who experienced viral suppression between week 24 and year 4 of ART. Liver enzyme, creatinine, and cholesterol levels and hematologic parameters were assessed during the trial. Cystatin C, high-sensitivity C-reactive protein (hs-CRP), interleukin 6 (IL-6), d-dimer, and soluble CD14 (sCD14) were assayed from cryopreserved specimens. RESULTS: Ninety-nine children (52 on PI-based and 47 on NNRTI-based ART) met inclusion criteria. The median age at initiation of ART was 6.5 years (interquartile range [IQR], 3.7-13.4 years), and 22% were aged <3 years at ART initiation; 56% of the PI-treated children received lopinavir/ritonavir, and 70% of NNRTI-treated children received efavirenz initially. We found no evidence of significant clinical toxicity in either group; growth, liver, kidney, and hematologic parameters either remained unchanged or improved between baseline and year 4. Total cholesterol levels increased modestly, but no difference between the groups was found. IL-6 and hs-CRP levels decreased more after 4 years in the NNRTI-based ART group. The median change in IL-6 level was -0.35 pg/ml in the PI-based ART group and -1.0 in the NNRTI-based ART group (P = .05), and the median change in hs-CRP level was 0.25 µg/ml in the PI-based ART group and -0.95 µg/ml in the NNRTI-based ART group (P = .005). CONCLUSION: These results support the safety of prolonged ART use in HIV-infected children and suggest that suppressive NNRTI-based regimens can be associated with lower levels of systemic inflammation.