Abstract: To reduce muscle injuries for workers engaged in manual handling tasks, a lasso-driven handling assistance exoskeleton is proposed, addressing the limitations of low flexibility and poor human-robot coupling in existing exoskeletons.The exoskeleton employs motor-driven lasso reels to enable flexible deployment/retraction of the lasso, providing lifting assistance. Firstly, the motion intention of wearers is recognized based on multi-sensor data fusion. And a hierarchical control strategy is adopted for the exoskeleton, with finite motion state machine and hybrid force-position control serving as the upper-and lower-level control of the control system. Secondly, performance testing and assistance evaluation are conducted on the exoskeleton. Experimental results demonstrate 11.7% reduction in biceps brachii muscle activation level and 37.3% reduction in lumbar erector spinae activation level during the 10 kg load lifting/carrying. During the 10 kg load transportation at a walking speed of 0.8 m/s, biceps brachii muscle activation level is decreased by 31.9%. These results validate the effective assistance capability of the exoskeleton for both arm and lumbar regions.