Objective To explore the effects of robot assisted training （RAT） combined with neural mobilization （NM） training on the recovery of upper limb functions in stroke patients. Methods A total of 110 stroke patients who met the inclusion criteria were selected as the subjects and randomly divided into a control group （n = 28）， RAT group （n = 27）， NM group （n = 28）， and combination group （n = 27）. All patients underwent routine upper limb occupational therapy. Additionally， the patients in the RAT group were treated with upper limb rehabilitation robots， those in the NM group underwent neural mobilization for treatment， those in the combination group were managed with robot-assisted training for upper limb rehabilitation and neural mobilization. Before treatment and 4 weeks after treatment， the modified Ashworth scale （MAS）， Fugl-Meyer assessment upper extremity （FMA-UE）， functional test for the hemiplegic upper extremity Hong Kong version （FTHUE-HK）， and modified Barthel index （MBI） were used to assess the effects. The surface electromyographic signals of the biceps and triceps at the maximum isometric voluntary contraction （MIVC） position during elbow flexion and extension were measured， the integrated electromyographic values （iEMG） were recorded and the synergistic contraction rate （CR） was calculated. Results There was no statistically significant difference （P > 0.05） between the four groups in the general information and pre-treatment assessments of MAS， FMA-UE， FTHUE-HK， MBI， iEMG， and CR.After 4 weeks， significant improvements were observed in all indicators compared to the pre-treatment assessments （P < 0.05）， with the exception of the triceps brachii CR， biceps brachii CR， and elbow extension MIVC biceps brachii iEMG in the control group.Among the group comparisons， all indicators showed statistically significant differences in mean or distribution （P < 0.05）， except for MAS and triceps brachii CR. The RAT group， NM group， and combination group all demonstrated significant improvements compared to the control group （P < 0.05）. Notably， the combination group exhibited a greater degree of improvement than the RAT and NM groups. Conclusion RATcombined with NM can reduce upper limb muscle tone in stroke patients. This approacheffectively promotes the establishment of normal movement patterns， improve upper limb motor function， and enhance activities of daily living. This combination is effective and worthy of further clinical promotion and application.