Compared with equilibrium data in pure solvent, variations of solubility are frequently observed in the presence of dissolved impurities. It is also well-known that impurities can inhibit the crystallization processes and lead to supersaturation barriers below which the growth of crystals is canceled. However, many papers dealing with the inhibiting effects of impurities in solution crystallization are rather unclear about the appropriate way of expressing the supersaturation in impure media. Indeed, as suggested in the present work, the latter can be defined either with respect to the solubility in pure solvent or in reference to the solubility of the impure solvent. Setting the correct reference for computing supersaturation in impure solutions is obviously a key issue for understanding and modeling the dynamics of elementary crystallization phenomena (i.e., primary and secondary nucleation, crystal growth, etc.) The present case study aims at clarifying this point. Solubility data were acquired by in situ ATR FTIR measurements of the concentration of ammonium oxalate in pure water and in impure water containing various concentrations of nickel sulfate dissolved as impurity. Supersaturation thresholds were observed and analyzed, according to the theoretical framework of the Kubota-Mullin impurity adsorption model. The experimental thresholds are attributed to the solubility drifts resulting from the pollution of the pure solvent by impurities. Finally we propose that the impurity-dependent solubility should be considered as the appropriate reference for defining supersaturation.