This paper presents a development of a modified Smith predictor for integrator-plus-dead-time (IPDT) process control. An IPDT process plant is intentionally modelled by a first-order-plus-dead-time (FOPDT) model within the predictor in order to nullify the effect of disturbances. Results show that the IPDT – FOPDT Smith predictor can successfully suppress disturbance loading. Stability analyses have carried out with the Lyapunov criterion for generic cases. Bode plot analysis shows a good stability for an IPDT process control case study. To reduce the overshoot of a process response during the transient response for a set-point tracking, a fuzzy-logic-based method for set-point weight tuning of a PID controller was applied. Simulation results demonstrate that this new modified Smith predictor has superior robustness to process dead time modelling error and good performance for set-point tracking and disturbance rejection.