A validated database of 70 molecules known to undergo biotransformation by CYP2C9 was collated. The molecular alignment program ROCS was used with the query molecule flurbiprofen as a basis for predicting the correct active site orientation of the CYP2C9 database molecules. The quality of the results obtained was excellent, with 39 of the first 44 molecules (89%) sorted by ROCS combination score having alignments that accounted for the experimentally observed site of oxidation. Transposition of the first 39 correctly aligned molecules into the CYP2C9 active site yielded an average site of metabolism to iron heme distance of 5.21 A, in good agreement with previous experimental observations. Molecular docking studies were also undertaken, but the results were less successful than the ROCS-based alignment method, indicating that ligand-based approaches with chemical typing are important in the prediction of metabolism by CYP2C9.