This paper considers the impact of exposure to violent conflict in utero and after birth, on a number of maternal and child health inputs and outcomes, using the Maoist insurgency that affected Nepal between 1996 and 2006 as a case study. Conflict intensity is measured by the number of conflict deaths by district and month, during the whole conflict period. Maternal fixed-effects estimation allows me to control for-, and shed light on- selection into becoming pregnant and giving birth at times of more intense conflict. Exposure to conflict in the first few years of life has an adverse effect on long-term child nutritional status in families with more than one young child. However, exposure to conflict in utero has both scarring and selection effects. The probability of miscarriage increases, and scarring effects dominate only for exposure to conflict in the third trimester of pregnancy. A range of mechanisms is explored. Use of antenatal care, (medical) help with delivery, and immunization coverage do not appear to decrease when conflict intensifies, and there is no evidence of conflict-related short-term malnutrition. This suggests a role for maternal stress and a low-level, prolonged nutritional deficit.