We investigate the impact of obstetrician supervision, as opposed to midwife supervision, on the short-term health of low-risk newborns. We exploit a unique policy rule in the Netherlands that creates a large discontinuity in the probability of a low-risk birth being attended by an obstetrician at gestational week 37. Using a fuzzy regression discontinuity design, we consistently find no health benefits from obstetrician supervision, despite increased rates of neonatal intensive care unit admissions among births supervised by obstetricians. These results indicate potential cost savings from increased use of midwifery care for low-risk deliveries.