It has often been thought that maternal and infant mortality were closely linked and responded in a similar manner to the same social, economic, and medical determinants of mortality rates. If there were a close link it would be expected to be strongest between maternal and neonatal mortality, because the latter is usually ascribed to natal (or endogenous) factors. Conversely, the link between maternal mortality and post-neonatal mortality might be expected to be slight because post-neonatal deaths are usually due to environmental (or exogenous) factors. In this paper, which is based largely on the period 1900–60, it is shown that against expectation the links between maternal mortality and all components of infant mortality, including neonatal mortality, are very slight. The explanation can be found by examining the immediate causes of maternal and neonatal deaths. The advantage of placing these mortalities side by side and comparing their response to various determinants is that it brings to the fore features of maternal and infant mortality which are not obvious if they are examined in isolation.