The substrates of lactate dehydrogenase (pyruvate and lactate) as well as an inhibitor (oxamate) are characterized using ab initio methods employing the basis sets 6-31+G(d) and 6-31+G(d,p) at both RHF and MP2 levels. The first aim of these studies is to explore the conformational preferences of these molecules and to compare the minimum energy conformation with that found in crystallographic structures of oxamate bound in the enzyme active site. In the lowest energy conformer of pyruvate, the carboxyl group is rotated about 80° from the plane defined by the carbonyl group and methyl carbon atom, a preference dominated by repulsion between the carboxyl and carbonyl oxygens. A relatively small barrier of ∼2kcal/mol for rotation of the carboxyl group is calculated at the MP2/6-31+G(d,p) level. In contrast, both oxamate and lactate prefer a nearly planar orientation of the carboxyl group, which is more like the enzyme-bound conformation of oxamate. In these molecules, the repulsion between the oxygen atoms is counterbalanced by intramolecular hydrogen bonds, which are maximized in a near-planar conformation. The barrier to carboxyl rotation in oxamate is ∼4kcal/mol at the MP2/6-31+G(d,p) level, while that for lactate is much higher, about 10kcal/mol at the RHF/6-31+G(d) level. Thus, even though the minimum-energy lactate conformation is more similar to the enzyme-bound oxamate conformation than that for pyruvate, a steeper conformational energy surface indicates that at least as much energy would be required for the smaller torsional changes which occur on binding.A second aim of these studies is to evaluate whether the semiempirical methods AM1 and PM3 commonly used in QM/MM calculations of enzyme reactions adequately reproduce higher-level ab initio conformational results. The AM1 method is shown to be accurate for pyruvate and lower-energy regions of lactate, but less accurate for oxamate. PM3 tends to overestimate barriers to conformational changes for these molecules.