Maintaining track tension in tracked vehicles minimizes the excessive load on the tracks and prevents the peal-off of tracks from the road-wheel, and adequately guarantees the stable and improved driving of the tracked vehicles. However, the track tension cannot be easily measured due to the limitation in the sensor technology, harsh environment, etc. In this study, the track tension is estimated in realtime from the measurable signals of tracked vehicles and controlled based on a fuzzy logic controller. The proposed control system is implemented on tracked vehicles and its performance is evaluated under various driving conditions.