ObjectiveTo compare intramuscular oxytocin, Syntometrine® and carbetocin for prevention of postpartum haemorrhage after vaginal birth.DesignRandomised double‐blinded clinical trial.SettingSix hospitals in England.PopulationA total of 5929 normotensive women having a singleton vaginal birth.MethodsRandomisation when birth was imminent.Main outcome measuresPrimary: use of additional uterotonic agents. Secondary: weighed blood loss, transfusion, manual removal of placenta, adverse effects, quality of life.ResultsParticipants receiving additional uterotonics: 368 (19.5%) oxytocin, 298 (15.6%) Syntometrine and 364 (19.1%) carbetocin. When pairwise comparisons were made: women receiving carbetocin were significantly more likely to receive additional uterotonics than those receiving Syntometrine (odds ratio [OR] 1.28, 95% CI 1.08–1.51, P = 0.004); the difference between carbetocin and oxytocin was non‐significant (P = 0.78); Participants receiving Syntometrine were significantly less likely to receive additional uterotonics than those receiving oxytocin (OR 0.75, 95% CI 0.65–0.91, P = 0.002). Non‐inferiority between carbetocin and Syntometrine was not shown. Use of Syntometrine reduced non‐drug PPH treatments compared with oxytocin (OR 0.64, 95% CI 0.42–0.97) but not carbetocin (P = 0.64). Rates of PPH and blood transfusion were not different. Syntometrine was associated with an increase in maternal adverse effects and reduced ability of the mother to bond with her baby.ConclusionsNon‐inferiority of carbetocin to Syntometrine was not shown. Carbetocin is not significantly different to oxytocin for use of additional uterotonics. Use of Syntometrine reduced use of additional uterotonics and need for non‐drug PPH treatments compared with oxytocin. Increased maternal adverse effects are a disadvantage of Syntometrine.Tweetable abstractIM carbetocin does not reduce additional uterotonic use compared with IM Syntometrine or oxytocin.