Recent research efforts on modeling personal thermal comfort support the integration of personalized preferences in optimal building control and further implementation in real buildings. This paper presents the development and field implementation of personal preference-based thermal control (i.e., a controller that provides thermal conditions to satisfy personal preferences) in real offices, emphasizing the role of model predictive control (MPC) and low-cost local sensing. Probabilistic thermal preference profiles were developed from experiments collected in identical private offices with controllable VAV systems. A low-cost thermal sensing network and a MPC framework were integrated into a centralized building management and control system. Customized, preference-based HVAC control was then experimentally implemented in the offices to (i) quantify the personal comfort penalty when using conventional wall thermostat versus local sensing-based operation for two distinct thermal preference profiles; (ii) evaluate the impact of personalized MPC (dynamic setpoint) on energy use and personal comfort compared with personalized simple feedback control (static setpoint), using local sensing; (iii) compare the personalized MPC performance for two distinct thermal preference profiles under different weather conditions. The results indicate the comfort benefits of monitoring local thermal conditions (vs wall thermostats) for different preference profiles and showed 28–35% energy savings with personalized MPC (vs personalized static setpoint control). The overall personalized MPC performance (and energy consumption) depends on the personal thermal preference characteristics and outdoor conditions.