Child protection is considered an appropriate government responsibility, but interventions into the family are also some of the most consequential for states. This study examines the normative basis for limiting parents’ freedom by exploring public attitudes about a child’s safety in the context of increasing risk. Using a randomized survey, we test the causal relationship between levels of risk and parental restrictions on representative samples in Norway and CA, US (n = 2148) – different welfare state and child protection models. Findings suggest that the public supports restricting parental freedom under conditions of risk and that severity of risk is taken into consideration. A majority favour restricting parental freedom under conditions of risk to the child; a minority resist restricting freedom, regardless of risk, and about one-third to one-half of respondents favour temporarily suspending parents’ rights by separating children to foster care. Residents of Norway are half as likely to support unrestricted parenting, regardless of risk, and are 1.5 as likely to endorse restricted parenting. Norwegians are also 20% less likely to support separating a child from his parent compared to US respondents. The study has implications for system design based on popular notions about parents’ freedom and family privacy.