Control systems tend to lose performance over time if their responses are not monitored and thus there is no support information on to how to make adjustments on them. Reliable controllers have complementary systems to identify and diagnose reductions in performance and also to implement predetermined solutions vis-à-vis the desirable type of output. The goal of this work was to analyze controller performance monitoring and causes diagnosis methods based in two indexes: historical benchmark and model based performance measurement. These methods were applied to situations of degraded performance simulated in the predictive control of a hydrotreating reactor, aiming the identification of the reduction in the controller performance and the discrimination of its causes. The obtained results can also be extended to several other chemical processes, once that the investigated process presents first order with dead-time dynamics, typical of these processes.