A two-phase approach to the control and operation of complex chemical processes at their optimum operating conditions is presented. The first phase consists of on-line parameter identification and state estimation of approximate nonlinear dynamic process models using on-line and off-line measurements. In the second phase, the optimum operating strategy is determined by integrating and optimizing this identified process model over a selected time horizon into the future. The method is particularly suited to those processes that exhibit slow dynamic responses and are subject to disturbances that have a significant economic impact. Examples include batch chemical reactors, large distillation towers, and processes with significant holdup times such as large fluidized-bed reactors.