A micro linear series elastic actuator (LSEA) and its mechatronic system were designed for force controlled robotic adaptive grasping. To enhance the measurement accuracy of the force control system, nonlinearity calibration experiment was performed to calibrate the nonlinearity of sensing system of the micro linear series elastic actuator by utilizing Bezier calibration method (BCM). An actuating force and displacement synchronous sensing method, which is based on dual displacement sensors, was proposed. Identification experiments were performed to develop the target deformation trajectory and actuating force observation models. A PID control strategy based on the target deformation trajectory model was established for actuating force control. Step force control and adaptive grasping experiments were performed to obtain the optimal parameters of the force controller and study the adaptive grasping characteristics of the series elastic actuated grasping system. Experimental results revealed that the micro LSEA could exhibit an accurate sensing and control ability without a force sensor. The LSEA had an extremely low overshoot in actuating force control. For the actuacting force with a amplitude of 15N, the percent overshoot was only 0.6%. In the force controlled robotic adaptive grasping experiment, the grasping posture of the fingertip could be adjusted by the actuating force control. As a result, the direction of the grasping force could be optimized to point to the object center of mass and enhance the grasping stability.