Intrinsic characteristics of distillation such as dead time and high nonlinearities do not allow the complete elimination of transient times when any external disturbance or set‐point change occurs. Thus, aiming at the use of easy‐tuning systems, a distributed‐action control in trays of a diabatic distillation unit with Smith's predictor was implemented in the Simulink environment to further reduce transient times and out‐of‐specification product. The distributed‐action strategy with Smith's predictor led to a reduction of 33.3 min (33 %) in the transient time of the top temperature control loop and 66 % in out‐of‐specification product, when compared with the conventional strategy, and thus is shown to be an efficient approach to increasing the productivity of distillation plants.