Background Differences in gait between people with multiple sclerosis (PwMS) and controls are well documented and effective in distinguishing between disability levels based on the Expanded Disability Severity Scale (EDSS) during preferred walking (PrW). However, clinically measured PrW does not consider higher cognitive demand required to perform more complex motor tasks in everyday living (e.g., conversing while walking). A dual-task walk (DTW) mimics this cognitive load, thus offering greater ecological validity when distinguishing between disability levels in PwMS. Purpose To evaluate the utility of spatio-temporal gait parameters during a DTW to distinguish between mild and moderate disability in PwMS. Methods Gait velocity, stride velocity, step width, step length and mean gait variability index were measured in a convenience sample of 44 PwMS (36=mild MS: EDSS 0 to 3.5, 7=moderate MS: EDSS 4.0 to 6.0) during PrW and DTW using a Protokinetic Zeno Walkway. Subjects were dichotomized into mild or moderate disability groups. Variable selection was performed using stepwise linear regression, including all gait metrics and covariates (sex, age, height, type of multiple sclerosis: relapsing or progressive, and disease duration). Logistic regressions were performed separately for PrW and DTW. Results During PrW, greater step length (cm) was associated with mild disability (R^2=0.39, ChiSquare=14.72, p<0.001). During DTW, both stride width (ChiSquare=4.69, p=0.03) and step length (ChiSquare=5.16, p=0.02) were associated with disability (R^2=0.38): people with moderate disability demonstrated increased step width and decreased step length. Discussion Results highlight the need to investigate differences in gait during DTW in addition to PrW in PwMS with different disability levels. However, generalizability is limited due to disproportionate sample sizes of disability groups. Still, DTW protocols may provide supplemental information for monitoring disability among PwMS, which can inform clinicians of the effectiveness of gait interventions or disease progression. Differences in gait between people with multiple sclerosis (PwMS) and controls are well documented and effective in distinguishing between disability levels based on the Expanded Disability Severity Scale (EDSS) during preferred walking (PrW). However, clinically measured PrW does not consider higher cognitive demand required to perform more complex motor tasks in everyday living (e.g., conversing while walking). A dual-task walk (DTW) mimics this cognitive load, thus offering greater ecological validity when distinguishing between disability levels in PwMS. To evaluate the utility of spatio-temporal gait parameters during a DTW to distinguish between mild and moderate disability in PwMS. Gait velocity, stride velocity, step width, step length and mean gait variability index were measured in a convenience sample of 44 PwMS (36=mild MS: EDSS 0 to 3.5, 7=moderate MS: EDSS 4.0 to 6.0) during PrW and DTW using a Protokinetic Zeno Walkway. Subjects were dichotomized into mild or moderate disability groups. Variable selection was performed using stepwise linear regression, including all gait metrics and covariates (sex, age, height, type of multiple sclerosis: relapsing or progressive, and disease duration). Logistic regressions were performed separately for PrW and DTW. During PrW, greater step length (cm) was associated with mild disability (R^2=0.39, ChiSquare=14.72, p<0.001). During DTW, both stride width (ChiSquare=4.69, p=0.03) and step length (ChiSquare=5.16, p=0.02) were associated with disability (R^2=0.38): people with moderate disability demonstrated increased step width and decreased step length. Results highlight the need to investigate differences in gait during DTW in addition to PrW in PwMS with different disability levels. However, generalizability is limited due to disproportionate sample sizes of disability groups. Still, DTW protocols may provide supplemental information for monitoring disability among PwMS, which can inform clinicians of the effectiveness of gait interventions or disease progression.