This study investigates and optimizes the performance of a localized dynamic personalized ventilation (PV) coupled with a mixing ventilation system (MV). A transient computational fluid dynamics model (CFD) was used to assess the velocity, temperature and CO2 fields around the occupant microclimate. The CFD model was also coupled with a transient bio-heat model to compute the segmental skin temperatures, overall thermal sensation and comfort. The model was then validated experimentally by using a thermal manikin in a climatic chamber representing an occupant in a typical office space. There was a good agreement between experimental and predicted values. The model was then used to optimize the operating frequency and average flowrate for the best thermal comfort and ventilation effectiveness. The optimal operating conditions of the localized dynamic airflow for a background temperature of 26 degrees C and a PV temperature of 22 degrees C, were for an average supply flowrate of 7.5 L/s and a frequency of 0.94 Hz providing an overall comfort of 0.95 (comfortable) on a comfort scale from very uncomfortable at 4 to very comfortable at +4 and a ventilation effectiveness of 77%. These conditions were able to provide the best compromise between comfort and air quality. They were also able to reduce energy costs by 21.34% compared to a constant personalized ventilation system providing the same levels of comfort. (C) 2017 Elsevier B.V. All rights reserved.