Upper limb motor dysfunction is a common complication after stroke, which has a negative impact on the daily life of the patients. Focal vibration (FV) has been used to improve upper limb motor function in acute and chronic stroke patients, but its application in subacute stroke patients has not been extensively explored. Therefore, the purpose of this study was to explore the therapeutic effect of FV on upper limb motor function in subacute stroke patients and its underlying electrophysiological mechanism. Twenty-nine patients were enrolled and randomized into two groups: a control group and a vibration group. The control group were treated with conventional therapy including passive and active physical activity training, standing and sitting balance exercises, muscle strength training, hand extension and grasping exercises. The vibration group were given conventional rehabilitation and vibration therapy. A deep muscle stimulator (DMS) with a frequency of 60 Hz and an amplitude of 6 mm was used to provide vibration stimulation, which was sequentially applied along the biceps muscle to the flexor radialis of the affected limb for 10 min, once a day, and 6 times a week. Both groups received treatments for 4 consecutive weeks. In the vibration group, the motor evoked potential (MEP) latency and the somatosensory evoked potential (SEP) latency were significantly shortened (P < 0.05) immediately and 30 min after vibration; the SEP amplitude and MEP amplitude were significantly increased (P < 0.05) immediately and 30 min after vibration. The MEP latency (P = 0.001) and SEP N20 latency (P = 0.001) were shortened, and the MEP amplitude (P = 0.011) and SEP N20 amplitude (P = 0.017) were significantly increased after 4 weeks in the vibration group. After 4 consecutive weeks, the vibration group showed significant improvements in Modified Ashworth Scale (MAS) (P = 0.037), Brunnstrom stage for upper extremity (BS-UE) (P = 0.020), Fugl-Meyer assessment for upper extremity (FMA-UE) (P = 0.029), Modified Barthel Index (MBI) (P = 0.024), and SEP N20 (P = 0.046) compared to the control group. The Brunnstrom stage for hand (BS-H) (P = 0.451) did not show significant differences between the two groups. This study showed that FV was effective in improving upper limb motor function in subacute stroke patients. The underlying mechanism of FV may be that it enhances the efficacy of sensory pathways and induces plastic changes in the sensorimotor cortex.Copyright © 2023. Published by Elsevier B.V.