Complex mathematical models necessitate large number of state variables to describe their dynamical behavior and performing numerical simulations on these models is challenging. This communication provides an efficient method of deriving reduced fractional order system from the higher order transfer function using JAYA method of optimization. Using direct synthesis approach, Smith predictor control strategy employing reduced fraction order model is suggested. Efficacy of the suggested model reduction method and control techniques is highlighted through numerical simulations conducted on the two case studies. The performance measures like settling time, ISE, IAE and overshoots are calculated to justify the superiority of the present work.