The surface water quality highly depends on the natural conditions and on the industrial, agricultural and other anthropogenic activities in the river catchment areas. The quality of the surface water is the key factor in the water quality management. The monitoring of the water quality indicators is requested to follow up (in time and space) the changes in the water quality, providing healthy water for the people. The water quality indicators (physical-chemical specific parameters such as: pH, electrical conductivity, dissolved oxygen, turbidity etc.) provide information about the pollution degree of different waters (stemming from wastewater treatment plants, precipitation, sewer lines, tributaries, etc.). The main aim of the research was to demonstrate the efficiency of artificial intelligence techniques for the water quality prediction. Consequently, the considered system was modelled using a neuro-evolutive technique, combining artificial neural networks (ANN) with differential evolution (DE) algorithm. The neural network acts as a system model, while the differential evolution algorithm is the optimizer with the function to determine the optimal parameters of the model. The simulation results pointed out that the approach is suitable to set up adequate models for the three parameters, pH, electrical conductivity, dissolved oxygen, respectively. For the fourth parameter (turbidity), an average absolute error of 30% was obtained, which indicates a more complex relation in rapport to the other parameters of the system. The DE-ANN approach can, thus, be further used in decision making solutions for water quality management. (C) 2018 Published by Elsevier B.V. on behalf of Institution of Chemical Engineers.