The recovery of independent walking is an important goal in stroke rehabilitation. The objective of this systematic review was to identify all outcome measures used in the stroke research literature that included an evaluation of walking ability and evaluate the concepts contained in these measures with reference to the International Classification of Functioning, Disability and Health (ICF) framework. Studies were selected for analysis if they used one or more standardized outcome measure incorporating an aspect of walking defined by the ICF. The outcome measure had to have published psychometric properties and specifically measure walking rather than mobility. The content of each outcome measure was classified with reference to the ICF subcategories for walking. Three hundred and fifty-seven studies met the selection criteria. Sixty-one different outcome measures were used a total of 848 times to measure walking ability. Six of the outcome measures reflected impairment and 52 reflected limitations of activity and participation. The 3 remaining outcome measures showed overlap between domains, reflecting aspects of both impairment and limitations in activity and participation. The three most frequently used measures (self-paced gait speed measured over a short distance, spatiotemporal parameters and fast gait speed) were used 350 times but only assessed one ICF subcategory. The Rivermead Mobility Index and the Adapted Patient Evaluation Conference System assessed the greatest number of ICF subcategories but were used only 19 times and once respectively. The authors concluded that the most frequently used outcome measures reflect only one aspect of walking ability: walking short distances. Furthermore, they observed that mobility tasks related to function in the community like walking long distances, around obstacles and over uneven ground, and moving around outside or in buildings other then the home are not well represented by outcome measures used in most studies.