Introduction: The Motricity Index (MI) can predict motor function after rehabilitation, but its minimal clinically important difference (MCID) has not been established. The primary study aim was to estimate the MCID value of the MI arm score. Methods: Between 2017 and 2018, 173 participants hospitalized with confirmed ischemic stroke were recruited into an observational rehabilitation study. Participants with motor weakness as measured by the Fugl-Meyer upper-extremity (FM-UE) and MI with complete baseline and follow-up assessments at 3 months were included in this analysis. The longitudinal recovery of the MI arm score was anchored to having a poor outcome based on the FM-UE recovery (<9) longitudinally. Results reported include the area-under-curve (AUC), along with sensitivity, specificity, and optimal cut-points based on maximizing the Youden statistic. Results: Sixty-nine patients (median [IQR] age 70  years; 48% male; 54% white) were included in the final analysis. Mean ± standard deviation outcome scores at 3-months were: MI arm: 83.19 ± 22.80; FM-UE: 53.04 ± 17.26. For the primary results, the MI arm score optimal MCID cutoff for observed recovery was 13 points with a sensitivity of 80% (95% Confidence Interval (CI)(67.6%, 92.4%)) and a specificity of 69.0% (95% CI (52.1, 85.8%)), and the AUC was 0.8082 (0.7007, 0.9157). Conclusions: This was the first study to report the MCID of the MI arm score, as anchored to the FM-UE recovery between acute evaluation and 3-months. The estimated optimal MCID of improvement in the MI arm score was 13 points.