Examination of the characteristics and locations of sleeping sites helps to document the social and ecological pressures acting on animals. We investigated sleeping tree choice for four groups of Colobus vellerosus, an arboreal folivore, on 298 nights at the Boabeng-Fiema Monkey Sanctuary, Ghana using five non-mutually exclusive hypotheses: predation avoidance, access to food, range and resource defense, thermoregulation, and a null hypothesis of random selection. C. vellerosus utilized 31 tree species as sleeping sites and the species used differed per group depending on their availability. Groups used multiple sleeping sites and minimized their travel costs by selecting trees near feeding areas. The percentage that a food species was fed upon annually was correlated with the use of that species as a sleeping tree. Ninety percent of the sleeping trees were in a phenophase with colobus food items. Entire groups slept in non-food trees on only one night. These data strongly support the access to food hypothesis. Range and resource defense was also important to sleeping site choice. Groups slept in exclusively used areas of their home range more often than expected, but when other groups were spotted on the edge of the core area, focal groups approached the intruders, behaved aggressively, and slept close to them, seemingly to prevent an incursion into their core range. However, by sleeping high in the canopy, in large, emergent trees with dense foliage, positioning themselves away from the main trunk on medium-sized branches, and by showing low rates of site reuse, C. vellerosus also appeared to be avoiding predation in their sleeping site choices. Groups left their sleep sites later after cooler nights but did not show behavioral thermoregulation, such as huddling. This study suggests that access to food, range and resource defense, and predation avoidance were more important considerations in sleeping site selection than thermoregulation for ursine colobus.