BACKGROUND: Retained placenta is associated with post-partum haemorrhage. Meta-analysis has suggested that umbilical injection of oxytocin could increase placental expulsion without the need for a surgeon or anaesthetic. We assessed the effect of high-dose umbilical vein oxytocin as a treatment for retained placenta. METHODS: In this double-blind, placebo-controlled trial, haemodynamically stable women with a retained placenta for more than 30 min were recruited from 13 sites in the UK, Uganda, and Pakistan. 577 women were randomly assigned by a computer-generated randomisation list stratified by centre to 30 mL saline containing either 50 IU oxytocin (n=292) or 5 mL water (n=285), which was injected into the placenta through an umbilical vein catheter. All trial participants, study workers, and data handlers were masked to individual allocations. The primary outcome was the need for manual removal of the placenta. Analysis was by intention to treat. This study is registered, number ISRCTN 13204258. FINDINGS: The primary outcome was recorded for all participants. We detected no difference between the groups in the need for manual removal of placenta (oxytocin 179/292 [61.3%] vs placebo 177/285 [62.1%]; relative risk 0.98, 95% CI 0.87-1.12; p=0.84). The need for manual removal was higher in the UK (overall 250/361 [69%]) than in Uganda (90/190 [47%]) or Pakistan (16/26 [62%]). Adverse events did not differ between the two groups. INTERPRETATION: Umbilical oxytocin has no clinically significant effect on the need for manual removal for women with retained placenta. FUNDING: WHO, WellBeing of Women, Pakistan Higher Education Commission.