Oxidative damage by non-steroidal anti-inflammatory drugs (NSAIDs) has been considered relevant to the occurrence of gastro-intestinal side-effects. In the case of chiral arylpropionate derivatives like ketoprofen (KPF), this mechanism has been evidenced for the R-enantiomer, especially when chiral inversion was observed, and lets us suppose the involvement of CoA conjugates. Glucose-6-phosphate dehydrogenase (G6PD) is the crucial enzyme to regenerate the GSH pool and maintain the intracellular redox potential. This enzyme is known to be down-regulated by palmitoyl-CoA thioester. We hypothesised then that G6PD is the target of carboxylic NSAIDs, via their CoA metabolites. We used molecular docking to localise a putative site in the human G6PD then we chose the Yeast orthologue, as the most suitable species to study experimentally the precise molecular interaction. KPF-CoA was effectively shown to bind covalently to the unique cysteine residue of the yeast enzyme. Binding was found to occur in the same site as palmitoyl-CoA. It was decreased in the presence of an allosteric inhibitor of G6PD, phospho(enol)pyruvate, and was not detected with G6PD of Leuconostoc mesenteroides, which does not possess the allosteric site. This site is distinct from the catalytic site, and probably allosteric, explaining the observed non-competitive inhibition of its activity by KPF-CoA. KPF-CoA was shown to induce the production of reactive oxygen species in Caco-2 cells, where its inhibition of G6PD activity was observed.