Movement of animals is a key process affecting population dynamics. Information on factors that affect pathway use is essential for identifying and protecting pathways, and important for maintaining connectivity among populations. We present an innovative, non-invasive, approach for predicting pathways of reintroduced Asiatic wild ass (Equus hemionus) in Israel, which is based on understanding the effects of landscape factors on pathways use. The approach includes: Predicting pathways, by employing a least cost pathway (LCP) GIS models based on several landscape factors, so as to efficiently direct a field survey and explore the wild ass’s general preferences of pathway types; Collecting empirical data by surveying the dung density of wild ass along each of the predicted pathways and using the data as an index of pathway use; Evaluating the predicted pathways against the empirical data collected, to estimate the general pathway preferences of the wild ass; and Developing and evaluating alternative generalized linear models, according to a priori hypotheses based on empirical data so as to quantify the effect of different landscape factors on pathway use. The analyses were conducted for the entire landscape, and then for two distinct landscape types, open landscape and landscape-barriers (mountain ridges), as subsets of the entire landscape. There were clear differences in the mean number of faeces counts between the LCPs, indicating that the wild ass prefers certain pathway types as a function of landscape features. We further found that the factors affecting E. hemionus pathway usage—vegetation; slopes; canyons; and 4-wheel drive routes—varied largely between the two major landscape types studied, demonstrating the importance of studying space use patterns at different landscape terrains. This information can be applicable to landscape planning measures that aim to enhance protection of the species. This approach provides a framework for studying animal space-use patterns of a variety of species, including elusive species, in a heterogeneous landscape.