In this paper, first we analyze why the conventional PID controller is not suitable to control integrating processes mathematically and intuitively. To overcome its structural limitation, we add an internal feedback loop, which can convert the integrating process to an open-loop stable process. Second, we propose a new and simple on-line process identification and autotuning method for the integrating process. Here we use-only the process input and output to obtain a process transfer function directly, while other previous closed-loop identification methods require the information of the closed-loop control structure. This identification method can be applied not only to integrating processes but also to all those encountered in chemical engineering processes which are open-loop stable or unstable. From simulation studies, we can recognize that, in spite of its simplicity, the proposed on-line process identification and autotuning method shows not only good model accuracy but also superior control performances for both servo and regulatory problems. It also gives strong robustness to measurement noises and disturbances.