OBJECTIVES: To assess whether intraoperative intravascular volume optimisation improves outcome and shortens hospital stay after repair of proximal femoral fracture. DESIGN: Prospective, randomised controlled trial comparing conventional intraoperative fluid management with repeated colloid fluid challenges monitored by oesophageal Doppler ultrasonography to maintain maximal stroke volume throughout the operative period. SETTING: Teaching hospital, London. SUBJECTS: 40 patients undergoing repair of proximal femoral fracture under general anaesthesia. INTERVENTIONS: Patients were randomly assigned to receive either conventional intraoperative fluid management (control patients) or additional repeated colloid fluid challenges with oesophageal Doppler ultrasonography used to maintain maximal stroke volume throughout the operative period (protocol patients). MAIN OUTCOME MEASURES: Time declared medically fit for hospital discharge, duration of hospital stay (in acute bed; in acute plus long stay bed), mortality, perioperative haemodynamic changes. RESULTS: Intraoperative intravascular fluid loading produced significantly greater changes in stroke volume (median 15 ml (95% confidence interval 10 to 21 ml)) and cardiac output (1.2 l/min (0.1 to 2.3 l/min)) than in the conventionally managed group (-5 ml (-10 to 1 ml) and -0.4 l/min (-1.0 to 0.2 l/min)) (P < 0.001 and P < 0.05, respectively). One protocol patient and two control patients died in hospital. In the survivors, postoperative recovery was significantly faster in the protocol patients, with shorter times to being declared medically fit for discharge (median 10 (9 to 15) days v 15 (11 to 40) days, P < 0.05) and a 39% reduction in hospital stay (12 (8 to 13) days v 20 (10 to 61) days, P < 0.05). CONCLUSIONS: Proximal femoral fracture repair constitutes surgery in a high risk population. Intraoperative intravascular volume loading to optimal stroke volume resulted in a more rapid postoperative recovery and a significantly reduced hospital stay.