Does single motherhood adversely affect infant health? This question is not easy to answer because of the endogeneity of coresidence during pregnancy. In this paper, we exploit quasi-natural variation in single motherhood from the moment of conception to that of birth arising from marriageable age restrictions and the reform of the laws in Korea. The Korean birth certificate dataset is unique in that it allows us to distinguish coresidence and legal marital status and further to identify the duration of pregnancy period without a partner. Results show that although coresidence with the partner during pregnancy is seemingly beneficial for infant health, it is mostly driven by selection into coresidence. Further, we do not find any significant advantage of legal marriage among young mothers.