Purpose
Generic preference-based measures often lack validity in states of impaired mobility. Our aim was to derive a novel health state classification system from the MobQoL tool, for the purpose of preference-based measurement of states of impaired mobility.

Materials and methods
Data were collected through online and postal surveys. Respondents were recruited through the NHS, charitable organisations and HealthWise Wales. Statistical and psychometric analyses were used to assess the validity and reliability of the MobQoL tool. Exploratory factor analysis and Rasch analysis were used to determine dimensional structure and to select items for the MobQoL health state classification system.

Results
Three hundred and forty-two respondents completed the survey. Respondents had a wide range of different mobility impairments. Nine of the MobQoL items demonstrated adequate validity and reliability. Exploratory factor analysis and Rasch analysis confirmed two sub-scales within the item structure: 1) physical and role functioning, and 2) mental wellbeing. Seven items were found to have adequate model fit and were retained in the final health state classification system, called the MobQoL-7D.

Conclusions
The MobQoL-7D contains seven dimensions of mobility-related quality of life: accessibility, contribution, pain/discomfort, independence, self-esteem, mood/emotions and anxiety. Population level preference weights are now needed for different states of mobility impairment.