Northern Botswana is one of the giraffe strongholds across its geographic range in the absence of, or with low impact from, the major anthropological threats faced elsewhere. Yet despite its conservation significance, until now, no giraffe specific ecological or behavioural studies have been undertaken. Moreover, Africa’s giraffe population has been significantly reduced over the last two decades, and the pressure on giraffe habitats and populations is likely to increase as the human population continues to expand, and the effects of climate change take their toll. As such, it is important to establish an ecological and behavioural baseline for giraffe in the unique ecosystems. This study provides baseline data on the behaviour and ecology of giraffe in the dry savannah and woodlands of the northern Chobe region and the wetland system of the Okavango Delta (NG26) in northern Botswana. The study first examines the home ranges, seasonal ranges and daily movements of giraffe in the two study areas. Next, the focus is on giraffe behaviours and activity budgets, and the effect of site, sex, season and time of day on behaviour. Lastly, the study describes the giraffe social grouping patterns in Chobe and examines non-random associations and spatial overlap as possible factors driving the population’s underlying fission-fusion system. Ecological and behavioural similarities were observed between the study areas, but also vast differences reflecting the adaptations made by giraffe in response to the unique set of environmental factors they face. Home ranges and daily movements were larger in Chobe where forage is more limited and patchily distributed, and ranges were larger during the dry season. Habitat, season, and sex were all found to be influential factors contributing to the observed variation in giraffe activity budgets. Social analyses revealed a complex social organisation whereby non-random associations and spatial overlap are drivers of a structured social network found within a fluid fission-fusion social system. Variation in pairwise association strengths and ranging patterns between the sexes suggest that males and females socialise differently. Association strength was generally greater for females than males though both sexes appeared to have preferred and avoided associates, indicating non-random groupings of individuals. Pairwise association strengths appear to be influenced more by social preferences and avoidances than spatial overlap, and shared space use has a greater influence over female social groupings.