In this paper, the problem of improving the performance of a discharge air temperature (DAT) system using a PID controller and augmenting it with neural network based tuning and tracking functions is explored. The DAT system is modeled as a SISO (single input single output) system. The architecture of the real time neuro-PID controller and simulation results obtained under realistic operating conditions are presented. The neural network assisted PID tuning method is simple to implement. Results show that the network assisted PID controller is able to track both constant and variable set point trajectories efficiently in the presence of disturbances acting on the DAT system. (C) 2003 Elsevier Ltd. All rights reserved.