Objectives: To investigate (1) whether previously observed changes in gait parameters in individuals with multiple sclerosis (MS) are the result of slower preferred walking speeds or reflect adaptations independent of gait speed; and (2) the changes in spatiotemporal features of the unstable swing phase of gait in people with MS. Design: Cross-sectional study assessing changes in gait parameters during preferred, slow (0.6m/s), medium (1.0m/s), and fast (1.4m/s) walking speeds. Setting: Gait laboratory with instrumented walkway and motion capture system. Participants: MS group with mild to moderate impairment (n_19, 16 women) with a median Expanded Disability Status Scale score of 3.75 (range, 2.5– 6), and a sex- and age-matched control group (n_19). Interventions: Not applicable. Main Outcome Measures: Gait speed, stride length, stride width, cadence, dual support time, swing time, and timing of swing foot and body/head center of mass during swing phase. Results: Individuals with MS walked at slower preferred speeds with longer dual support times compared with controls. In fixed-speed conditions, dual support times were longer and swing times were shorter in MS compared with controls. Stride width was wider for all speed conditions in the MS group. In fixed-speed conditions, the MS group positioned their head and body centers of mass closer to the anterior base of support boundary when entering the unstable equilibrium of the swing phase. Conclusions: Longer dual support time is part of a gait strategy in MS that is apparent even when controlling for the confounding effect of slower preferred speed. However, a gait strategy featuring longer dual support times may have limitations if potentially destabilizing swing dynamics exist, which especially occur at walking speeds other than preferred for people with MS.