Background: Impaired mobility is associated with functional dependence, frailty, and mortality in prevalent patients undergoing dialysis. We investigated risk factors for mobility impairment, (poor gait speed) in patients incident to dialysis, and changes in gait speed over time in a 2-year longitudinal study. Methods: One hundred eighty-three patients enrolled within 6 months of dialysis initiation were followed up 6, 12, and 24 months later. Grip strength, health-related quality of life, and comorbidities were assessed at baseline. Outcomes were (a) baseline gait speed and (b) change in gait speed over time. Gait speed was assessed by 4-meter walk. Multivariate linear regression was used to identify risk factors for low gait speed at baseline. For longitudinal analyses, linear mixed effects modeling with gait speed modeled over time was used as the outcome. Results: Participants were 54.7 ± 12.8 years old, 52.5% men, 73.9% black with mean dialysis vintage of 100.1 ± 46.9 days and median gait speed 0.78 (0.64-0.094) m/s. Lower health utility and grip strength, diabetic nephropathy, and walking aids were associated with lower baseline gait speed. Loss of 0.1 m/s gait speed occurred in 24% of subjects at 1 year. In multivariate mixed effects models, only age, walking aid use, lower health utility, and lower handgrip strength were significantly associated with gait speed loss. Conclusions: In our cohort of incident dialysis patients, overall gait speed is very low and 54.2% of the subjects continue to lose gait speed over 2 years. Older age, lower handgrip strength, and quality of life are risk factors for slowness. Patients at highest risk of poor gait speed can be identified at dialysis initiation to allow targeted implementation of therapeutic options.