In this paper, a nonlinear autoregressive with exogenous input (NARX) network was utilized to model a conveyor-belt grain dryer using a set of input-output data which was collected during an experiment to dry paddy grains. The resulting NARX model has achieved a remarkable modelling accuracy compared to other previously reported modelling techniques. To control the considered dryer, a simplified Type-2 adaptive neuro-fuzzy inference system (ANFIS) controller was proposed. The effectiveness of this controller was demonstrated by several performance tests conducted by computer simulations. Moreover, a comparative study with other related controllers further confirmed the superiority of the proposed dryer controller.