Abstract Objective: To evaluate patterns of successful gait strategy in Huntington disease (HD) at various stages of illness to improve fall avoidance and maintenance of independence. Design: Repeated measurements of gait kinematic parameters and joint performance during gait cycles of six HD patients compared to 30 age-matched controls. Setting: A standard gait laboratory. Subjects: Six HD patients, rank-ordered for disease severity from minimal chorea to generalized dystonia, selected because they were ambulatory despite 3 to 17 years' disease duration. One patient was from a nursing home (walked with assistance) and five were living independently, either alone or with a working spouse who was the caregiver. Main Outcome Measures: Standardized gait evaluations (retroreflective markers on standard bony landmarks) from five video angles, fed into digitizer to computer-generate joint angles and standard gait kinematic parameters. Results: Wide variability in gait kinematic parameters and joint interaction plots (phase plane and angle-angle plots) was observed between individuals and successive trials of the same limb, tending to increase with disease severity. Joint interaction plots show that random, highly variable distractions from planned trajectories are more apparent distally. Conclusions: Chorea in HD does not appreciably affect the center of gravity during ambulation, and the consistency of gait profiles at heel strike shows that the ultimate target is achieved in each step despite random and frequent variability during the gait cycle.