A knowledge-based controller(KBC) used in process control systems is presented. It has three features: first, it does not need the mathematical model; secondly, the adjustable parameters of KBC have practical meanings, so they can be determined easily using human experience, and thirdly, the contribution of KBC to a controlled plant is separated into two parts: steady states contribution and transient contribution. A simple fuzzy reasoning is employed to tune the KBC parameters. The experimental and simulation results show that KBC is very effective especially when there are variations in the process dynamics.