Objective: to compare cost-effectiveness of two models of maternity service delivery: Midwifery Group Practice (MGP) at a birth centre and standard care (SC). Design: a prospective non-randomised trial. Setting: an Australian metropolitan hospital. Method: women at 36 weeks gestation were approached in the birth centre or hospital antetal clinics between March and December 2008. Of 170 consecutive women who met birth centre eligibility criteria, 70% (n챱9) were recruited to the study. Women (MGP n쵲 or standard care n쵰) were followed through to 6 weeks postpartum. Publically funded care costs were collected from women's diaries, handheld pregncy health records, medical records and the hospital accounting system. Main outcome measures: health-care costs to the hospital and government. Alysis: generalised linear models with covariates of age, nulliparity, private health insurance (yes/no) and household income category. Findings: women receiving MGP care were less likely to experience induction of labour, required fewer antetal visits, received more posttal care, and neotes were less likely to be admitted to special care nursery than those receiving standard care. Statistically signi?cant lower costs were found for women and babies receiving MGP care compared with women receiving standard care during pregncy, labour and birth and postpartum to 6 weeks. MGP resulted in lower costs for the hospital ($AUD4,696 vs. $AUD5,521 po0.001) and the government ($AUD4,722 vs. $AUD5,641 po0.001). When baby costs were excluded MGP care remained statistically signi?cantly cheaper than standard care. Conclusion: for women at low-risk of birth complications, Midwifery Group Practice was cost effective, and women experienced fewer obstetric interventions compared with standard maternity care. The evidence suggests Midwifery Group Practice is safe and economically viable.