Structure-activity relationship studies of valproic acid (VPA) derivatives have revealed a quantitative correlation between histone deacetylase (HDAC) inhibition and induction of neural tube defects (NTDs) in the NMRI-exencephaly-mouse model, but this correlation has been, so far, limited to congeners with a carboxylic acid function. Whereas the classical HDAC inhibitor trichostatin A is active only as a hydroxamate but not as a carboxylic acid, we found that neither VPA amides nor hydroxamates inhibit HDACs, but can cause NTDs; e.g., 2-pentyl-4-pentynoic hydroxamic acid with its S-enantiomer being the potent teratogen. We therefore investigated the hypothesis that hydroxamic acid derivatives of VPA might be metabolized in vivo and may possibly be pro-teratogenic, as had been shown for valpromide but not valproic hydroxamic acid. We developed two stereoselective quantification methods based on chiral derivatization of VPA hydroxamates with (1R,2S,5R)-(-)-menthylchloroformate and carboxylic acid derivatives with (S)-(-)-1-naphthylethylamine, followed by gas chromatography-nitrogen phosphor detector analysis of biological samples. We then determined the pharmacokinetic profiles of S-2-pentyl-4-pentynoic hydroxamic acid and of S-2-pentyl-4-pentynoic acid in mice. S-2-Pentyl-4-pentynoic hydroxamic acid was found to be extensively metabolized to the corresponding carboxylic acid without affecting the stereochemistry at position C2. Furthermore, the metabolite S-2-pentyl-4-pentynoic acid was found to be very stable in vivo, with an extended half-life of 4.2 h compared with that of VPA, 1.4 h. Comparison of the individual HDAC inhibition abilities of additional VPA amides and hydroxamates, as measured by cellular and enzymatic assays, led us to the conclusion that both classes of VPA derivatives can be pro-teratogenic.