The objective was to determine the cytochrome P450s (CYPs) responsible for the stereoselective and regiospecific hydroxylation of ketamine [(R,S)-Ket] to diastereomeric hydroxyketamines, (2S,6S;2R,6R)-HK (5a) and (2S,6R;2R,6S)-HK (5b) and norketamine [(R,S)-norKet] to hydroxynorketamines, (2S,6S;2R,6R)-HNK (4a), (2S,6R;2R,6S)-HNK (4b), (2S,5S;2R,5R)-HNK (4c), (2S,4S;2R,4R)-HNK (4d), (2S,4R;2R,4S)-HNK (4e), (2S,5R;2R,5S)-HNK (4f). The enantiomers of Ket and norKet were incubated with characterized human liver microsomes (HLMs) and expressed CYPs. Metabolites were identified and quantified using LC/MS/MS and apparent kinetic constants estimated using single-site Michaelis-Menten, Hill or substrate inhibition equation. 5a was predominantly formed from (S)-Ket by CYP2A6 and N-demethylated to 4a by CYP2B6. 5b was formed from (R)- and (S)-Ket by CYP3A4/3A5 and N-demethylated to 4b by multiple enzymes. norKet incubation produced 4a, 4c and 4f and minor amounts of 4d and 4e. CYP2A6 and CYP2B6 were the major enzymes responsible for the formation of 4a, 4d and 4f, and CYP3A4/3A5 for the formation of 4e. The 4b metabolite was not detected in the norKet incubates. 5a and 4b were detected in plasma samples from patients receiving (R,S)-Ket, indicating that 5a and 5b are significant Ket metabolites. Large variations in HNK concentrations were observed suggesting that pharmacogenetics and/or metabolic drug interactions may play a role in therapeutic response.