Abstract and Figures In this paper, the implementation of the controller based on neural networks for controlling two-tank system is presented. A ready-made mathematical model of the Amira DTS200 system, which is a typical example of a slow nonlinear process, is used. Among the most important applications of artificial neural networks is their application in the control of nonlinear processes. The applied controlling structure represents Direct Inverse Control. Experimental results of the obtained process response for a given reference input using implemented inverse controller are given.