An intelligent system based on an array non-specific chemichal sensors was developed. Detection was carried out using an array of potentiometric sensors based on PVC membranes of different selectivity. This work shows how the use of arrays of potetiometric sensors without selective response to specific analytes, when combined with modelling abilities of the A, is an interesting approach for the simultaneous determination of analytes together with its interferents. The proposed approach is implemented with an electronic toungue formed by three potentiometric sensors (Sodium, potassium, and cadmium ISEs) The measurement system comprised the detection system (sensor array incorporate into the system in series ) and a potentiometer built in our laboratories, enabling up to six electrodes to be controlled and which was itself controlled by computer with in-house software developed. The generated responses are processed with an adequately trained ANN, the availability of measuring systems not requiring of any stage of interference removal would facilitate largely the development of specially robust and compact system suited for environmental monitoring of this parameter. The diluted standard solutions needed for system learning. The data set was randomly divided into three data sets, calibration (60 samples ) prediction (15 samples) and test set (13 samples). The combined response was modelled by means of Artificial Neural Networks (A). In order to identify the ANN which provided the best model of the electrode responses. The networks parameters optimized simultaneously and its usefulness in determining of K + Na + Cd 2+ was then tested.The system was used for analysis of real samples with satisfactory result.