This study focuses on achieving both energy saving and thermal comfort in Activity-Based Working (ABW) offices. In ABW offices, workers move according to their work style, and have different thermal preferences. Therefore, we aim to minimize thermal dissatisfaction by intentionally creating areas with appropriate environmental deviations, allowing workers to autonomously choose their preferred environment. By integrating air conditioning energy reduction into these variations, we aimed to balance comfort with energy savings. The objective of this study is to develop a spatial com-fort evaluation model that considers worker mobility. A field demonstration was con-ducted in a Tokyo ABW office where a smartphone-based location detection system was installed. To understand employees’ thermal preferences, a survey was conducted and the temperature acceptable range by workers was defined. When the space temperature deviates from the worker’s acceptable range, a penalty score of the evaluation model is applied. Additionally, in this model, no penalty is given if the acceptable temperature falls within a worker’s defined accessible radius. The model evaluated 555 randomly generated temperature distributions using the location data and the temperature preferences of workers over a period of nine months. When worker movement was not considered and worker’s accessible radius was set to 0 m, a uniform thermal environment resulted in the smallest penalty. However, when the accessible radius was set to 10 and 20 m, the penalty was smaller in spaces with deviations. This suggests that using this new evaluation model contributes to improving worker comfort by encouraging movement under varying environments.