HIV infected individuals have poorer response to hepatitis B vaccine (HBV) compared to normal host. Intradermal administration (i.d.) facilitates the exposure of antigen to antigen-presenting cells compared to intramuscular administration (i.m.). HIV-infected children aged 1-18 years with CD4%≥15% or 200 cells/mm(3) who had negative HBs Ag, antiHBs, and antiHBc were randomized to receive 3-dose of HBV via i.d. (2 μg/dose) or i.m. (10 μg/dose) route at months 0, 2, and 6. AntiHBs titers were measured at months 2, 6 and 7 after first HBV. AntiHBs≥10 mIU/mL was considered protective and AntiHBs>100 mIU/mL was considered good response. Participants included 41 in i.d. and 39 in i.m. arms. 64% had completed 3-doses HBV during infancy. The mean (SD) of age, nadir CD4% and current CD4% were 12 (3.3) years, 10.6 (7.9)% and 28 (8.0)% respectively. 91% were on HAART and 84% had undetectable HIV-RNA. Proportion of children with protective antiHBs in i.d. vs. i.m. group were 19.5% vs. 25.6% at month 2, 56.1% vs. 76.9% at month 6, and 90.2% vs. 92.3% at month 7 (NS, all). The geometric mean (95% confidence interval) of antiHBs titer in i.d. vs. i.m. group were 112.5 (34.4-367.6) vs. 141.2 (49.4-404.1) mIU/mL at month 2 (p=0.74), 70.4 (39.8-124.4) vs. 132.1 (79.4-219.8) mIU/mL at month 6 (p=0.10), and 157.0 (103.0-239.3) vs. 458.9 (324.0-647.0) mIU/mL at month 7 (p<0.001). However, only 56.1% of the i.d. arm had good response to HBV compared to 82.1% in the i.m. arm (p=0.01). The predictors for being a good responder to HBV were i.m. administration [OR 4.0, 95%CI 1.4-11.8, p=0.012] and body weight <35 kg at baseline [OR 3.8, 95%CI 1.3-10.8, p=0.013]. No adverse events grade 3/4 occurred. In conclusion, HIV-infected children without severe immune suppression, both i.d. and i.m. routes of HBV resulted in similar rates of protective antibody titers. However, high antibody titers to HBV were more common with i.m.; therefore, i.m. administration is preferred.