Introduction: The de Morton Mobility Index (DEMMI) is a unidimensional clinical instrument proven reliable and valid for measuring mobility in the elderly, but its responsiveness has not been evaluated when applying the methodology recommended by COSMIN.Purpose: To evaluate the responsiveness and the predictive validity of the DEMMI.Materials and methods: Elderly people, referred for municipality-based rehabilitation, were consecutively included. Eight hypotheses regarding the relationship between the change scores in the DEMMI and other functional measures of mobility were tested by assessing the participants before and after rehabilitation. Three months after completing the rehabilitation, the participants' functional status was evaluated.Results: 250 elderly people (mean age 85 years, 63% women) were included. The change scores in the DEMMI and the other measures correlated (r = 0.422 to 0.547), but only three out of the eight hypotheses were confirmed. A DEMMI score of less than 48 and less than 67 can predict inability of ambulating close to the participant's residence and using public transport, respectively.Conclusions: We found a moderate responsiveness and a high predictive validity of the DEMMI score. The DEMMI appears to be an appropriate instrument for monitoring change in mobility and predicting functional status among elderly patients undergoing rehabilitation.Implications for rehabilitationThe de Morton Mobility Index (DEMMI) is a clinical instrument for measuring mobility.This study shows that the DEMMI was responsive to change in mobility and predicted disability among a diverse group of elderly individuals referred to rehabilitation after hospital discharge.The DEMMI is applicable as an outcome measure in trials investigating the effect of rehabilitation and for clinical decision-making concerning geriatric rehabilitation.