Six ungulate species inhabit the Colombian Guiana Shield, a patchily-distributed biogeographic region characterized by poor nutrient soils and by harboring the only forest in Colombia growing on a Precambrian surface. In Colombia, two protected areas are located on the Colombian Guiana Shield, the Tuparro National Natural Park (PNNT, for its Spanish initials) and the Puinawai National Natural Reserve (RNNP, for its Spanish initials), which have a large difference in the configuration and availability of native forest. These ungulates are three deer species (Odocoileus virginianus, Mazama americana and M. nemorivaga), two peccary species (Pecari tajacu and Tayassu pecari) and one tapir specie (Tapirus terrestris). The aims of this study were to estimate and compare the abundance of those six ungulate species in two protected areas located within the Guiana Shield region in Colombia and analyze the relationship with the landscape forest amount and distance to human settlements. The study sites were in PNNT where the landscape is dominated by natural savannas mixed with riparian forests, and RNNP covered by tropical forests. We estimated the relative abundance of each ungulate species using repeated presence-absence data with the occupancy model with heterogeneous detection of Royle and Nichols (2003). Presence detection was obtained from both direct observations and tracks. The direct sightings and tracks we sampled along six transects in PNNT and nine in RNNP (244.4 km at PNNT and 257 km at RNNP) and track plots of 0.25 m 2 (211 at PNNT and 297 at RNNP). The relative abundance of tapir, collared peccary (P. tajacu) and brocket deer (Mazama spp.) were different between protected areas; we found no differences in the white-lipped peccary abundance (T. pecari). Relative abundance estimation was higher at RNNP for tapirs and brocket deers. In contrast, collared peccary was more abundant in the PNNT. There was positive effect of floodplain forest area abundance of white-lipped peccary. The distance to human settlements had positive effect on collared peccary abundance and negative effect on tapir and white tailed deer (O. virginianus). The higher abundance in the RNNP for tapirs and Mazama spp., brocket deer was unrelated to the forest area, so the difference may be related to local habitat conditions. The white-lipped peccary is sensitive to the forest area in the landscape, but has also been found that riparian forests act as natural corridors. Failure to find differences in abundance can be an indicator of the high habitat quality for these species. The collared peccary has been positively associated to a moderate degree of subdivision of forest at landscape (number of patchs), which could explain the higher abundance in the PNNT. The positive relationship of white-lipped peccary with the floodplain forest coincides with the patterns observed in other Amazonian areas. Finally, the negative effect of the distance from settlements for brocket deer and tapir may be related to low hunting pressure on these species, con...