Fossils from Sibrambang and Djambu, two sites in the Padang Highlands, were collected by Dubois in the late 1880s. These collections, alongside the deposits from Lida Ajer, have for over 100 years been our only insights into the Pleistocene mammalian history of Sumatra. Despite their importance, their chronological context has remained elusive. Here, we provide the first direct dates of fossils from Sibrambang and Djambu. Uranium-thorium series dating indicates that it is likely that the fossils from Djambu were derived from at least three periods: (1) >500 ka (beyond or close to the limit of the applied dating technique); (2) close to 85 ka (but not younger); and (3) close to 38 ka (but not younger). Sibrambang, too, has a mix of fossils with different ages, and it is hard to say how many distinct time intervals may be present. Conservatively, there are at least two:(1) >149 ka; and (2) >55 ka (but not younger than that). Stable carbon and oxygen isotope analyses of fossils from both sites indicate largely rainforest conditions during this period, except for one elephant specimen (>500 ka), which is reconstructed here as a mixed feeder. These data, combined with previous studies, hint at more open environments in Sumatra during (periods of ) the Middle Pleistocene, although significantly more data will be required to confirm this. Our results have implications for previous palaeoecological analyses involving these sites, as well as for the taxonomy of fossil orangutan (Pongo).