Objectives: this study aimed to investigate the construct validity and responsiveness of performance-based and self-reported measures of strength, mobility and balance after hip fracture.Design: secondary analysis of clinical trial data.Subjects: a total of 148 older people undergoing hip fracture rehabilitation.Methods: correlation coefficients assessed construct validity. Internal responsiveness was assessed by calculating effect sizes (ES) I and II. Area under the receiver operating characteristic curve (AUC) assessed external responsiveness with change in EuroQol as the reference.Results: correlations between performance-based and self-reported measures were small to medium (strength r = 0.17, mobility r = 0.45 and balance r = 0.37). The most responsive performance-based measures included walking speed (ESI 1.7, ESII 1.2), Physical Performance and Mobility Examination (ESI 1.3, ESII 1.0) and chair-rise test (ESI 1.1, ESII 0.8). Self-reported mobility (ESI 0.8, ESII 0.6) and strength (ESI 0.8, ESII 0.6) were more responsive than self-reported balance (ESI 0.3, ESII 0.2). External responsiveness (AUC) was greatest for walking speed (0.72) and lowest for the measures of body sway (0.53).Conclusion: self-reported and performance-based indices appear to assess different constructs and may provide complementary information about physical functioning in people after hip fracture. Measures of strength and mobility showed greater ability to detect change than measures of balance. © The Author 2012. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved.