Africa’s conservation areas have vast road networks as a result of motorized vehicle game viewing. In conservation areas these roads provide increased accessibility to previously inaccessible areas, often with negative ecological impacts, such as wildlife fatalities and decreased habitat quality and connectivity. Additionally, tar roads often have a greater negative effect on mammal species than dirt roads, altering their spatial distribution and behaviour to a greater extent. The Kruger National Park, South Africa, has 2294 km of public roads within the park, of which 850 km is tarred – carrying a higher traffic density than dirt roads – and 1444 km is dirt. The aim of this study was to compare the ecological impacts of tar and dirt roads on medium to large mammal species in the Kruger National Park, hypothesizing that the tar roads within the park have greater negative impacts on the park’s mammal species. I focused on three response variables: 1. observation likelihoods; 2. species group sizes and 3. distances to the road, comparing tar to dirt roads. Sampling was done at 1 km intervals, for 430 km of tar and 430 km of dirt roads, in savanna vegetation during the Austral winter of 2017 and 2019. In total 476 sampling points were on tar roads (401 systematic and 75 ad hoc) and 451 on dirt roads (369 systematic and 82 ad hoc). At each sampling point, all observed medium to large mammals, their group size, distance of the closest individual to the road, GPS coordinates, traffic volume and percentage cloud cover were recorded. Chi-square analyses were used to identify associations between species presence and road type. To determine the relationship between group size and road type, as well as the observed distance from the road and road type, Generalised Linear Models (GLMs) were used. The results refute the hypothesis of increased ecological impacts associated with tar roads, as the majority of the commonly occurring mammal species were not disproportionately associated with a specific road type. My results indicated that, as a tourist, one has an equal likelihood of observing the majority of the common mammal species from either road type. The results suggest that the increased traffic intensity of tar roads does not decrease one’s chances of viewing game, as hypothesized. The commonly occurring species were also observed at similar distances from both road types, and all were observed on the road, suggesting a limited barrier effect, if any. Previous studies on the impacts of road type on mammals both support and contradict my results. Elephant (Loxodonta africana) was the only species that was associated with a road type, although the complete opposite was found to what was hypothesized. Elephant occurred more often alongside tar roads (albeit only in the north of the park) than dirt roads, in larger groups next to tar roads and closer to tar roads. Potential reasons that are discussed are higher quality forage, plant species richness, group vigilance and predator avoidance. My results can have an influence on park management decisions and communication with tourists. Tourists can be informed that they do not need to make use of dirt roads, which are often less accessible, to view game. In terms of road management, my results suggest that the temporary closure of dirt roads for rehabilitation should not affect game viewing. It is, however, suggested that further research is needed to compare the influence of public to private road areas on animal ecology within the Kruger National Park, as well as to compare the influence of road areas to completely natural areas. / Mini Dissertation (MSc (Environmental Ecology))--University of Pretoria, 2020. / Thuthuka programme - National Research Foundation (NRF) of South Africa (Grant no. 94103). / Zoology and Entomology / MSc (Environmental Ecology) / Restricted