Background: Two CYP2B6 variants, 516G→T and 983C→T, define high efavirenz (EFV) concentration strata (slow metabolizers). Contributions by CYP2A6, UGT2B7, CYP3A4 and CYP3A5 variants have been reported, but have been inconsistent. We characterized associations between EFV concentrations and genetic variants in 84 AIDS Clinical Trials Group (ACTG) study participants, all with CYP2B6 516/983 slow metabolizer genotypes. Methodology: Several randomized ACTG treatment trials (esp. A5202, A5095/A5097s, and ACTG 384) included EFV-containing arms, and plasma sampling for EFV at multiple study weeks. We included subjects with minimal intraindividual variability between ≥2 EFV determinations at 8-19 hours post-dose. Slow metabolizer CYP2B6 genotypes (CYP2B6 516TT, 516T/983C, and 983CC) were available from previous analyses. Genotyping of UGT2B7 was by sequencing, and 198 additional loci (107 polymorphic) in 40 drug metabolism/transport genes by iPLEX ADME PGx platform. Race/ethnicity was by self-report. Associations with the mean of log10-transformed EFV concentrations were assessed by linear regression. Results: Analyses included Black (56%), White (29%), and Hispanic (15%) subjects. CYP2B6 genotypes were 516TT in 71 subjects, 516T/983C in 12, and 983CC in 1. EFV concentrations were somewhat higher with 983C alleles (P=0.12). In univariate analyses (additive models), CYP2A6 -48T→G (rs28399433) was significantly associated with increased EFV concentrations in all subjects (P=3.8x10E-4, Figure, top); this was consistent in Black and White groups analyzed separately. UGT2B7 735A→G (rs28365062, *1c) and 801T→A (rs7438284, *2) were not associated with EFV concentrations by univariate analysis (p=0.34 and P=0.63), and after controlling for CYP2A6 -48T→G (p=0.42 and P=0.67, respectively). Homozygosity for UGT2B7 735GG (n=3), but not heterozygosity, was associated with higher EFV concentrations (P=0.006, Figure, bottom). With correction for multiple comparisons, no iPLEX ADME PGx variant other than CYP2A6 -48T→G was associated with EFV concentrations, including 22 in CYP2A6 and 3A4/5. Conclusions: In CYP2B6 slow metabolizers, CYP2A6 -48T→G is associated with higher EFV concentrations. An effect of UGT2B7 735A→G on EFV concentrations, if true, is much less. No other significant associations in these or other genes were found. These findings support a contribution of CYP2A6, and possibly UGT2B7 variants to even higher EFV concentrations in CYP2B6 slow metabolizers.