The effect of mineral salt on the reactive dye removal was evaluated by using flat ultrafiltration (UF) membranes made of polyethersulfone (PES) and regenerated cellulose (C). Five reactive dyes varied in molecular weight (Reactive Orange 16, Remazol Brilliant Blue R, Reactive Orange 20, Reactive Black 5, Reactive Read 120) were used in the UF experiments. The applied membranes were characterized by the cut-off value equal to 10 kDa. The ultrafiltration process was performed in a dead-end mode under the transmembrane pressure of 0.2 MPa. Dye concentration in model solutions was equal to 100 mg/dm 3 . Two various mineral salts (NaCl and Na 2 SO 4 ) were chosen to simulate the composition of dye-house baths. The concentration of monovalent and divalent salt in model dye solutions amounted to 1–2 g/dm 3 . Transport and separation properties of the UF membranes towards reactive dyes and mineral salts were evaluated. It was found that the separation efficiency of all tested dyes was dependent on the mineral salt presence. The effect of mineral salt on dye retention was more pronounced for membrane made of regenerated cellulose (C) than for polyethersulfone membrane (PES). In the case of the C membrane the dye retention coefficients varied in the wide range – from 55 to 83%, whereas for PES membrane the degree of dye separation was less diverse and amounted to 81–96%. The presence of mineral salt in water and dye solutions generally worsen the membrane permeability due to osmotic pressure increase as well as enhancing adsorptive fouling of dye particles in membrane matrix.