Purpose : To develop a preliminary proposal for the revision of the severity code of the ICIDH-D 1980. Method : Quantitative analysis (polytomous Rasch analysis) of linked existing data sources including items about walking and dressing disability. Results : The Rasch analysis provided estimates of threshold parameters for walking and dressing item categories. Factor analysis showed that more than one dimension was present, but that the first factor could definitely be interpreted as 'disability'. The reliability of the solutions was satisfactory (0.88 for walking and 0.91 for dressing). Conclusion : Based on the results, tentatively a new severity code is proposed with more distinct categories of 'difficulty ' than the existing code, and the use of technical aids and personal assistance in the higher ranks. The Rasch method could be a useful tool for calibrating and measuring disability, as well as for converting existing disability data into a new uniform severity of disability code.