Quality of recovery after anaesthesia is an important measure of the early postoperative health status of patients. We attempted to develop a valid, reliable and responsive measure of quality of recovery after anaesthesia and surgery. We studied 160 patients and asked them to rate postoperative recovery using three methods: a 100-mm visual analogue scale (VAS), a nine-item questionnaire and a 50-item questionnaire; the questionnaires were repeated later on the same day. From these results, we developed a 40-item questionnaire as a measure of quality of recovery (QoR-40; maximum score 200). We found good convergent validity between QoR-40 and VAS (r = 0.68, P < 0.001). Construct validity was supported by a negative correlation with duration of hospital stay (rho = -0.24, P < 0.001) and a lower mean QoR-40 score in women (162 (SD 26)) compared with men (173 (17)) (P = 0.002). There was also good test-retest reliability (intra-class ri = 0.92, P < 0.001), internal consistency (Cronbach's alpha = 0.93, P < 0.001) and split-half coefficient (alpha = 0.83, P < 0.001). The standardized response mean, a measure of responsiveness, was 0.65. The QoR-40 was completed in less than 6.3 (4.9) min. We believe that the QoR-40 is a good objective measure of quality of recovery after anaesthesia and surgery. It would be a useful end-point in perioperative clinical studies.