Background. When people with stroke recover gait speed, they report improved function and reduced disability. However, the minimal amount of change in gait speed that is clinically meaningful and associated with an important difference in function for people poststroke has not been determined. Objective. The purpose of this study was to determine the minimal clinically important difference (MCID) for comfortable gait speed (CGS) associated with an improvement in the modified Rankin Scale (mRS) score for people between 20 to 60 days poststroke. Design. This was a prospective, longitudinal, cohort study. Methods. The participants in this study were 283 people with first-time stroke prospectively enrolled in the ongoing Locomotor Experience Applied Post Stroke (LEAPS) multi-site randomized clinical trial. Comfortable gait speed was measured and mRS scores were obtained at 20 and 60 days poststroke. Improvement of ≥1 on the mRS was used to detect meaningful change in disability level. Results. Mean (SD) CGS was 0.18 (0.16) m/s at 20 days and 0.39 (0.22) m/s at 60 days poststroke. Among all participants, 47.3% experienced an improvement in disability level ≥1. The MCID was estimated as an improvement in CGS of 0.16 m/s anchored to the mRS. Limitations. Because the mRS is not a gait-specific measure of disability, the estimated MCID for CGS was only 73.9% sensitive and 57.0% specific for detecting improvement in mRS scores. Conclusions. We estimate that the MCID for gait speed among patients with subacute stroke and severe gait speed impairments is 0.16 m/s. Patients with subacute stroke who increase gait speed ≥0.16 m/s are more likely to experience a meaningful improvement in disability level than those who do not. Clinicians can use this reference value to develop goals and interpret progress in patients with subacute stroke.