This contribution is aimed at providing a survey of the limitations of the peak-height technique. It is shown that the shape and the slope of calibration curves give a precise insight into the separation mechanism and can warn the analyst against improper selection of background electrolyte or sample dilution. This work investigates three typical situations where the sample contains (i) the minor analyte only, (ii) the minor analyte and a nonstacking bulk component, (iii) the minor analyte and a stacking bulk component, and various modes of measurement of the calibration curves differing in the medium used for the dilution of the original sample (water, background electrolyte, solution of the bulk component). Based on simple theoretical models, the shapes of the calibration curves are derived and elucidated for all modes, clearly demonstrating that only a few modes can provide useful results while most of them seem useless for quantitation. The knowledge of regularities affecting the character of calibration curves can be used for optimization of the analysis so that reliable results and highest attainable sensitivity can be reached. The outcome of this study brings a clear instruction how to successfully quantitate trace analytes even in samples with a complex and variable matrix(*).