The Etosha Giraffe Research project in Namibia focuses on the social relationships among giraffes, which have previously been described as having little social structure nor strong bonds between individuals (Dagg and Foster, 1976; Le Pendu et al., 2000; Moss, 1976). Giraffes live in a fission-fusion society characterised by frequent changes of associates, with males adopting a roaming strategy to search for widely distributed female groups (Bashaw et al., 2007; Bercovitch et al., 2006; Dagg and Foster, 1976; Shorrocks and Croft, 2009). Female giraffes come into oestrus for only 4 days every two weeks after giving birth, with a pregnancy phase of ~15 months and births occurring throughout the year (Bercovitch et al., 2006; Dagg and Foster, 1976). Weak maternal bonds between mothers and calves have been reported, to the extent that it is difficult to establish which individual is the mother of a calf (Dagg and Foster, 1976). As a consequence of weak cow-calf bonds, juveniles may be left in a crèche group during the day (i.e. day-care) in the care of one or two adult females. It is possible that these crèche groups may consist of related juveniles and that bonds between juveniles and older females may persist through time (Bashaw et al., 2007). These ideas are yet to be tested in wild populations, however some evidence points towards the existence of a matrilineal social structure, since subadult females have been recorded associating with their mothers in a population of giraffes in the Namib Desert, while males exhibit greater dispersal and are thus likely to leave female groups at an early age (Brand, 2007; Fennessy, 2004; van der Jeugd and Prins, 2000). Recent research that has focussed on relationships among captive female giraffes has shown that giraffes show stronger preferences to be with older relatives (Bashaw et al., 2007), suggesting that adult female giraffes who preferentially associate with each other may be close relatives. Thus, earlier beliefs about the lack of social structure in giraffes may be incorrect.