Context: Antiretroviral therapy (ART) for pediatric human immunodeficiency vir us (HIV) infection has evolved from simple nucleoside reverse transcriptase inhi bitor (NRTI)-regimens to complex combination therapies based largely on evidenc e from clinical trials. However, the integration of novel ART into the clinical care of pediatric HIV infection has not been examined. Objectives: To describe c hanges in the treatment of pediatric HIV infection in the United States from 198 7-2003, to assess concordance of initial regimens with US pediatric guidelines, and to identify predictors of the first regimen switch. Design, Setting, and Pa rticipants: The study population included 766 perinatally HIV-infected children in the Pediatric AIDS Clinical Trials Group 219C cohort born before January 1, 2004, who had not participated in an ART clinical trial at 219C enrollment or du ring follow-up. Main Outcome Measures: Proportion of children receiving specifi c ART regimens, proportion of children initiating ART according to pediatric gui delines, and time to first regimen switch (risk of switching). Results Single an d dual NRTI regimens were used most frequently through 1997. In 1998, 2 years af ter protease inhibitors were approved for adult HIV infection and at the time pe diatric guidelines were issued, regimens of highly active antiretroviral therapy including a protease inhibitor became most frequently used. From 1998-2003, 22 %of children initiated ART with a regimen not recommended by pediatric guidelines. In multivariate regression, the risk of switching decreased with age at ART ini tiation (hazard ratio [HR], 0.96; 95%confidence interval [CI], 0.94-0.99) and increased with year of initiation (HR, 1.28; 95%CI, 1.23-1.33). The risk of sw itching was higher in children who started with 1 NRTI (HR, 8.05; 95%CI, 5.80- 11.18), 2 NRTIs (HR, 4.08; 95%CI, 3.08-5.40), or an unconventional regimen (HR , 6.23; 95%CI, 3.36-11.54) vs children who started with a protease inhibitor- containing regimen; and in children who initiated ART at CD4 T lymphocyte percen tages less than 15 vs 15 or greater (HR, 2.90; 95%CI, 1.03-8.13). Conclusions: There was a short lag between the identification of novel ART and its adoption in the pediatric community. A variety of regimens were used, including some unor thodox therapies. Important predictors of first regimen switch were identified. Context: Antiretroviral therapy (ART) for pediatric human immunodeficiency vir us (HIV) infection has evolved from simple nucleoside reverse transcriptase in hi bitor (NRTI) -regimens to complex combination therapies based largely on evidenc e from clinical trials. However, the integration of novel ART into the clinical care of pediatric HIV infection has not been examined. Objectives: To describe c hanges in the treatment of pediatric HIV infection in the United States from 198 7-2003, to assess concordance of initial regimens with US pediatric guidelines, and to identify, predictors of the first regimen switch. Design, Setting, and Pa rticipants: The study population included 766 perinatally HIV-infected children in the Pediatric AIDS Clinical Trials Group 219C cohort born before January 1, 2004, who had not participated in an ART clinical trial at 219C enrollment or du ring follow-up. Main Outcome Measures: Proportion of children receiving specifi c ART regimens, proportion of children Results Single an d dual NRTI regimens were used most frequently through 1997. In 1998, 2 years af ter protease inhibitors were approved for adult HIV infection and at the time pe diatric guidelines were issued, regimens of highly active antiretroviral therapy including a protease inhibitor became most frequently used. From 1998-2003, 22% of children initiated ART with a regimen not recommended by pediatric guidelines. In multivariate regression, the risk of switching decreased with age at ART iniation (hazard ratio [HR], 0.96; 95% confidence interval [CI], 0.94-0.99) and increased with year of initiation (HR, 1.28; 95% CI, 1.23-1.33). The risk of sw itching was higher in children who started with 1 NRTI (HR, 8.05; 95% CI, 5.80-11.18), 2 NRTIs (HR, 4.08; 95% CI, 3.08-5.40), or an unconventional regimen , 6.23; 95% CI, 3.36-11.54) vs children who started with a protease inhibitor- containing regimen; and in children who ART ART at CD4 T lymphocyte percen tages less than 15 vs 15 or greater (HR, 2.90; 95% CI, 1.03-8.13). Conclusions: There was a short lag between the identification of novel ART and its adoption in the pediatric community. A variety of regimens were used, including some unor thodox therapies. Important predictors of first regimen switch were identified.