Recovery from motor paralysis is facilitated by affected patients' recognition of the need for and practice of their own exercise goals. Neurorehabilitation has been proposed and used for the treatment of motor paralysis in stroke, and its effect has been verified. If an expected score for the neurorehabilitation effect can be calculated using the Fugl-Meyer Motor Assessment (FMA), a global assessment index, before neurorehabilitation, such a score will be useful for optimizing the treatment application criteria and for setting a goal to enhance the treatment effect. Therefore, this study verified whether the responsiveness to a treatment method, the NovEl intervention using repetitive transcranial magnetic stimulation and occupational therapy (NEURO), in patients with post-stroke upper extremity (UE) motor paralysis could be predicted by the pretreatment FMA score. No control group was established in this study for NEURO treatment. To analyze the recovery of the motor function in the UE, delta-FMA was calculated from the pre-and post-FMA scores obtained during NEURO treatment. The probability of three levels of treatment responsiveness was evaluated in association with delta-FMA score (<5, 5≤ delta-FMA <10, and ≥10 as non-responders; responders; and hyper-responders, respectively) according to the reported minimal clinically important difference (MCID). The association of the initial FMA scores with post-FMA scores, from the status of the treatment responsiveness, was determined by multinomial logistic regression analysis. Finally, 1,254 patients with stroke, stratified by FMA scores were analyzed. About 45% of the patients who had FMA scores ranging from 30 to 40 before treatment showed improvement over the MCID by NEURO treatment (odds ratio = 0.93, 95% CI = 0.92-0.95). Furthermore, more than 25% of the patients with more severe initial values, ranging from 26 to 30, improved beyond the MCID calculated in the acute phase Hamaguchi et al. Motor Recovery Predictions in Stroke (odds ratio = 0.87, 95% CI = 0.85-0.89). These results suggest that the evaluated motor function score of the UE before NEURO treatment can be used to estimate the possibility of a patient recovering beyond MCID in the chronic phase. This study provided clinical data to estimate the effect of NEURO treatment by the pretreatment FMA-UE score.