PID controllers have become the most popular control strategy in indiistrial processes due to tlie versatility and tumiiiig capabilities. The iiicorporation of aiito-tunning tools have increased the use of this kind of controllers. In this paper we i)ropose a neural network-based self-tunning scherne for on-line updating of PID parameters, which is based on integ'ral error criteria (lAE, ISE, ITAE, ITSE).