Darwin used observations of infants as evidence for his evolutionary hypotheses about human agency, in three ways. First, human actions that appear fully formed at the start of life, like sucking, were deemed reflexes or instinctive fruits of evolution. Second, infant actions show in a clear and simple form the foundations of human agency. Third, when there is no direct way of proving how complex forms of human action evolved, their growth in infancy provides a working model for natural, simple-to-complex development that is analogous to evolution. Two texts exploit these arguments: Expression (1872) and ‘A Biographical Sketch of an Infant’ (1877). The former concentrates on crying and weeping. The latter focuses on some of the distinctively human forms of agency described in Descent. A key omission in the evidence Darwin’s infant observations provide for his theory is a test of infants’ capacity for group-interaction. Evidence from such a test is critical to acceptance of Descent’s thesis that adaptations to group-life ground the most distinctive forms of human behaviour. Only recently have scientists sought this evidence. From these we know that preverbal infants do have a capacity for ‘groupness.’ Darwin’s observations of young children show a robustness and prescience borne out by contemporary research.