Industrial textile processing comprises the operations of pretreatment, dyeing, printing and finishing. These production processes are not only heavy consumers of energy and water; they also produce a substantial amount of chemical pollution. Of all dyed textile fibres, cotton occupies the number-one position, and more than 50% of its production is dyed with reactive dyes, owing to their technical characteristics. Unfortunately, this class of dyes is also the most unfavorable one from the ecological point of view, as the effluents produced are relatively heavily colored, contain high concentrations of salt and exhibit high BOD/COD values. Dyeing 1 kg of cotton with reactive dyes requires an average of 70-150L water, 0.6 kg NaCl and 40 g reactive dye. The composition of the dye bath which we propose to treat contains solid particles (cotton fibres), dyeing auxiliaries (organic compounds), hydrolyzed reactive dyes, substantial quantities of alkalis (sodium carbonate and soda ash) and very high concentration of sodium chloride or sodium sulfate. This paper presented the state of the art of the different processes currently used for the treatment of dye house wastewaters and evaluated a four-step process [1] to recover the water and the mineral salts, while leaving the spent dyes in the reject stream. Processes evaluated included (1) cartridge filtration to remove textile fibres, (2) acidification to make the brine recovered, suitable for reuse and further dyeing operations, (3) nanofiltration (NF) to concentrate the hydrolyzed dyes and (4) reverse osmosis (RO) to further concentrate the salts for reuse in the dyeing process. A cut-off of 100 mu m is sufficient to trap textile fibres, regardless of the type of effluent and the texture of the textile dyed. The hydrolyzed reactive dyes present in the treated effluents comprise the entire range of possible types of reactive dyes. For this acidification, we studied the influence of the concentration of sodium chloride, the influence of the temperature and we verified that the volume neither depends on the concentrations of reactive hydrolyzed dyes nor sodium chloride. After defining the nanofiltration membrane, we studied the effect of the pH, temperature, pressure and velocity as well as the experimental procedure on the permeate flux, recovery of the salt and removal of the color. An increase of either of the parameters temperature and pressure leads to an increase of the permeate flux. On the other hand, a rise in the pH leads to a decrease of the permeate fiux. The retention factor of the sodium chloride is low when the concentration of sodium chloride is high in the retentate. Our aim was to recover 80-90% of the sodium chloride, but our experiments showed that the recovery went as high as 99%. Depending on the dyes used, the experimental procedure can be carried out in one, two or three steps. The dye retention level was always higher than 98%. After studying the operating variables, experiments with the recycled brines in new dyeing operations were carried out with specimen dyeings prepared with usual water using different classes of reactive dyes. There was no difference in the results in terms of depth, shade or fastness properties, whichever type of water was used. These last results therefore validate our process and its special innovative feature: recycling not only the water but also the mineral salts. (c) 2005 Elsevier B.V. All rights reserved.