Despite universal access to health care, there is a significant health gradient in childhood in France, i.e. a positive association between family income and child health. This gradient can be interpreted in two ways, and parents' health plays a role in both of them. First, family income may have a positive impact on parents' health, which in turn may have a positive influence on child health. Second, parents' health may have an effect on both family income and child health, creating a misleading correlation between family income and child health. The impact of public policies on child health will be different depending on the interpretation of the gradient. If family income influences parents' health, which itself has an impact on child health, then a policy that increases the income of some households will improve parents' health, and then child health. But if the correlation between income and child health is misleading, then such a policy will have no impact on child health. However, if parents' health has a causal effect on child health, any policy that improves parents' health is also likely to benefit children. This article uses data from the 1994-2008 Enquête santé et protection sociale (ESPS) survey to study the relationship between family income and several aspects of child and parents' health. Our results suggest that income has an effect on child weight and digestive health, even when we take parents' health into account. This result points to the role of living conditions in social health inequalities during childhood.