OBJECTIVE To determine whether intra-umbilical vein injection with saline solution, with or without oxytocin, reduces the need for manual removal of placenta compared with expectant management. METHODS Multicenter, randomised controlled trial. METHODS Eleven hospitals in four cities of Argentina: Buenos Aires, Corrientes, Rosario, and Salta. METHODS Two hundred and ninety-one women showing no evidence of placental separation thirty minutes after vaginal delivery. METHODS Three different management strategies: 1. intra-umbilical vein injection of saline solution plus oxytocin; 2. intra-umbilical vein injection of saline solution alone; and 3. expectant management. METHODS Primary: manual removal of the placenta. Secondary: blood loss after trial entry, haemoglobin level at 24 to 48 hours and at 40 to 45 days after delivery, blood transfusion, curettage, infection, and days of hospital stay. RESULTS Rates of subsequent manual removal were similar: intra-umbilical vein injection of saline solution plus oxytocin (58%; RR 0.92; 95% CI 0.73-1.15), or saline alone (63%; RR 1.00; 95% CI 0.80-1.24), compared with expectant management (63%). There were also no detectable effects of the active managements on any of the secondary measures of outcome. CONCLUSIONS Based on evidence available from randomised controlled trials, including this trial, it is unlikely that intra-umbilical injection with or without oxytocin, is clinically useful. We recommend that this intervention should not be used in third stage management of labour.