The sleeping habits of wild white-handed gibbons (Hylobates lar) were investigated to assess the risk of predation and predation-avoidance behavior. Sleeping sites were distributed throughout home ranges, including areas where they overlapped with neighbors, and appeared to be selected independently of habitat characteristics. Individuals did not build night nests or otherwise manipulate the vegetation around the sleeping place but slept on open branches. Group members usually slept in separate trees, and, except for females with infants, they never shared a sleeping place. Sleeping trees were entered several hours before dusk and were used for about 14-17 h. The majority of sleeping trees were used only once, and fewer were selected repeatedly by the same or other group members. Usually females with infants went into a sleeping tree first, then juveniles, and last were mostly subadult and adult males. Intragroup competition over access to a sleeping place was observed once. Average time difference between the first and last group member to enter a sleeping tree was 13 min. The sequence of departure from sleeping trees was more variable. Gibbon sleeping habits seem to primarily reflect adaptations to minimize predation risk. The predation-risk hypothesis was indirectly supported by observations of mobbing pythons, alarm calls given in response to birdes of prey flying low over the canopy, and more importantly by 1) the predominant use of large sleeping trees, which were among the tallest trees available, particularly by adult females with small infants and juveniles, 2) an unpredictable long-term pattern of reuse of sleeping places, and 3) inconspicuous presleep behavior. Am.