ABSTRACT: A large number of structural determinations of compounds containing 2-hy-droxy-3,5-di-nitro-benzoic acid (I) and its various deprotonated forms, 2-hy-droxy-3,5-di-nitro-benzoate (II) or 2-carb-oxy-4,6-di-nitro-phenolate (III), are biased. The reason for the bias follows from incorrectly applied constraints or restraints on the bridging hydrogen, which is involved in the intra-molecular hydrogen bond between the neighbouring carb-oxy-lic/carboxyl-ate and oxo/hy-droxy groups. This hydrogen bond belongs to the category of resonance-assisted hydrogen bonds. The present article suggests corrections for the following structure determinations that have been published in Acta Crystallographica: DUJZAK, JEVNAA, LUDFUL, NUQVEB, QIQJAD, SAFGUD, SEDKET, TIYZIM, TUJPEV, VABZIJ, WADXOR, YAXPOE [refcodes are taken from the Cambridge Structural Database [CSD; Groom et al. (2016 ?). Acta Cryst. B72, 171-179]. The structural features of the title mol-ecules in all the retrieved structures, together with structures that contain 3,5-di-nitro-2-oxidobenzoate (IV), are discussed. Attention is paid to the localization of the above-mentioned bridging hydrogen, which can be situated closer to the O atom of the carboxyl-ate/carb-oxy-lic group or that of the hy-droxy/oxo group. In some cases, it is disordered between the two O atoms. The position of the bridging hydrogen seems to be dependent on the pKa (base) although with exceptions. A stronger basicity enhances the probability of the presence of a phenolate (III). The present article examines the problem of the refinement of such a bridging hydrogen as well as that of the hydrogen atoms involved in the hy-droxy and primary and secondary amine groups. It appears that the best model, in many cases, is obtained by fixing the hydrogen-atom position found in the difference electron-density map while refining its isotropic displacement parameter.