Objective: To determine if gait speed or walking distance is a better predictor for community walking after stroke.
Methods: Data from the FIT-Stroke trial were used in a cross-sectional design. Community walking was measured with a self-administered questionnaire.