Objective To investigate the clinical efficacy of low-frequency transcranial magnetic stimulation (rTMS) on post-stroke upper limb spasticity and its mechanism. Methods From September, 2015 to December, 2017, 23 patients with post-stroke upper limb paralysis were randomly divided into control group (n=13) and experimental group (n=10). Both groups received routine rehabilitation, and the experimental group received 1 Hz rTMS at primary motor area (M1) for eight weeks. They were assessed with modified Ashworth Scale (MAS), modified Barthel Index (MBI) and Fugl-Meyer Assessment-Upper Extremities (FMA-UE) before and after treatment, while the activation under fMRI in the task state was observed and the laterality index (LI) was calculated. Results The scores of MAS, FMA-UE and MBI improved after treatment in both groups (Z>2.121, t=6.248, P<0.05), and improved more in the experimental group than in the control group (Z>2.084, t=-2.095, P<0.05). The ipsilateral M1, ipsilateral sensory motor cortex and bilateral supplementary motor area were activated more in the control group than in the experimental group during the movement of affected hand. LI in the M1 increased after treatment in both groups (Z>2.366, P<0.05), and was more in the experimental group than in the control group (Z=-2.430, P<0.05). There was a positive correlation between the change of LI in the M1 and the improvement of the MAS and FMA-UE (r>0.612, P<0.05). Conclusion Low-frequency rTMS may improve the motor function and spasticity of upper limb after stroke by promoting reorganization of the cortex and inducing normalization of cortical function.