Abstract This paper studies mental health spillovers among new mothers. We exploit variation in the mental health of peers in mother groups in the Danish public postnatal care program. We show that municipal nurses assign mothers arbitrarily to groups conditional on a narrow set of well-defined characteristics. Exposure to a depressed peer in the group increases mothers' mental health care uptake by 11 percent two years after birth. We document worse self-reported mental health and labor market outcomes for treated mothers. Exploring heterogeneity, we find suggestive evidence for mental health deterioration, rather than increased demand for health care, as mechanism