We use 35 years of administrative data to document how newborn health varies with the business cycle in Spain. In panel regressions that include province and year fixed effects as well as province trends, we show that children have significantly better health outcomes at birth in times of high unemployment: a 10 percentage-point increase in the unemployment rate is significantly associated with about 2 log-points higher birth-weight, almost 2 percentage points fewer babies with low birth weight, 0.6 points fewer babies with very low birth-weight, and a 0.4-point drop in mortality rates in the first 24 h We explore several potential mechanisms. First, we show that the documented association is not driven by in-utero selection: we do not find that high unemployment is associated with more miscarriages, abortions, or stillbirths. Second, we explore the role of composition in terms of parental characteristics. We find evidence that unmarried and younger parents, who typically have unhealthier babies, are relatively less likely to have children when unemployment is high. Finally, we show that there are fewer first births during recessions, and birth order is strongly positively correlated with health at birth. Birth order can account for up to one fifth of the countercyclicality of birth weight.