Objective: To assess the cost-effectiveness of treatment with nifedipine compared with atosiban in women with threatened preterm birth. Design: An economic analysis alongside a randomised clinical trial (the APOSTEL III study). Setting: Obstetric departments of 12 tertiary hospitals and seven secondary hospitals in the Netherlands and Belgium. Population: Women with threatened preterm birth between 25 and 34 weeks of gestation, randomised for tocolysis with either nifedipine or atosiban. Methods: We performed an economic analysis from a societal perspective. We estimated costs from randomisation until discharge. Analyses for singleton and multiple pregnancies were performed separately. The robustness of our findings was evaluated in sensitivity analyses. Main outcome measures: Mean costs and differences were calculated per woman treated with nifedipine or atosiban. Health outcomes were expressed as the prevalence of a composite of adverse perinatal outcomes. Results: Mean costs per patients were significantly lower in the nifedipine group [singleton pregnancies: €34,897 versus €43,376, mean difference (MD) −€8479 [95% confidence interval (CI) −€14,327 to −€2016)]; multiple pregnancies: €90,248 versus €102,292, MD −€12,044 (95% CI −€21,607 to € −1671). There was a non-significantly higher death rate in the nifedipine group. The difference in costs was mainly driven by a lower neonatal intensive care unit admission (NICU) rate in the nifedipine group. Conclusion: Treatment with nifedipine in women with threatened preterm birth results in lower costs when compared with treatment with atosiban. However, the safety of nifedipine warrants further investigation. Tweetable abstract: In women with threatened preterm birth, tocolysis using nifedipine results in lower costs when compared with atosiban.