Profilometry is established in erosion research. However, in the case of dentine, factors such as the demineralised organic matrix, desiccation effects, or type of measuring device may have an impact on the measurement results, which were investigated in the present study. Dentine specimens were eroded with citric acid (1%, pH 2.6) for 5, 10, 15, 20, 30, 60, 90, and 120 min (n = 15 each). For each specimen, tissue loss was determined under various conditions - before/after enzymatic matrix removal, under standardised wet and desiccated (2/10 min) conditions - with non-contact and contact profilometry. In the presence of matrix, under wet conditions, non-contact profilometry revealed almost no tissue loss. Values (mean ± SD) ranged between 0.3 ± 0.7 μm (5 min) and 3.4 ± 1.5 μm (120 min). Contact profilometry increased values significantly (range: 2.9 ± 1.1 to 30.6 ± 5.8 μm). Desiccation (2 min) significantly increased values, except for 5 min of demineralisation, for non-contact profilometry (range: 0.8 ± 1.3 to 22.1 ± 5.5 μm), and decreased values for contact profilometry up to 15 min and increased them as from 90 min (range: 0.9 ± 1.2 to 33.0 ± 5.5 μm); results after 10 min of desiccation were comparable. After the removal of matrix, under wet conditions, values were distinctly higher (non-contact: 3.5 ± 0.8-55.5 ± 7.4 μm; contact: 4.2 ± 1.3-57.8 ± 8.1 μm). Desiccation (10 min) lowered values by about 2-5 μm due to specimen deformation. Bland-Altman comparisons of various outcomes revealed distinct significant proportional and relative biases. Loss of mineralised tissue cannot be adequately quantified in the presence of matrix. Desiccation leads to matrix shrinkage and specimen deformation. Most importantly, tissue loss values obtained in the presence or absence of matrix are not proportional. Therefore, if mineral status is the target criterion, matrix removal and moisture control are prerequisites.