Objective: We aimed at comparing clinical/immunological outcomes in human immunodeficiency virus (HIV)-infected patients who were treated with CYP2B6-guided and conventional efavirenz (EFV) therapy., Methods: This study was a 24-week prospective randomized controlled trial. Eligible patients were HIV-infected adults yet to start antiretroviral therapy. Twenty-four HIV-infected patients were recruited and randomly assigned to genotype CYP2B6 polymorphism before ART initial dose. Patients with CYP2B6 *6/*6 received 400 mg EFV-based regimen and those with other genotypes received 600 mg EFV-based therapy., Results: For CYP2B6 polymorphism, 12 patients were extensive metabolizers, ten patients were intermediate metabolizers, and only two patients were poor metabolizers (*6/*6). The overall mean EFV plasma concentrations were similar in both groups. The mean drug concentrations (standard deviation) were 1.675 (0.963), 1.445 (0.778), and 1.899 (0.808) µg/mL at week 4, 12, and 24, respectively. The CYP2B6 *6/*6 patient who received low dose of EFV had lower mean EFV level than those who received a normal dose, 1.916 versus 3.915 µg/mL (P<0.001), respectively. Seventy percent of the patients had neuropsychiatric adverse events, especially dizziness., Discussion: There was a trend toward association of the CYP2B6 polymorphism and plasma EFV concentrations in this study. Reduced EFV dose should be considered in CYPB6 *6/*6 carrier to keep the drug concentration in therapeutic range.