Abstract We studied the habitat use of mountain nyala Tragelaphus buxtoni in the northern edge of the Bale Mountains National Park, Ethiopia. The aims of this study were to: (1) measure and quantify habitat-specific stem bite diameters of mountain nyala foraging on common natural plant species in two major habitat types (i.e. grasslands versus woodlands), and (2) quantify the bite rates (number of bites per minute) and the activity time budgets of mountain nyala as functions of habitat type and sex-age category. We randomly laid out three transects in each habitat type. Following each transect, through focal animal observations, we assessed and quantified stem diameters at point of browse (dpb), bite rates, and time budgets of mountain nyala in grasslands versus woodlands. Stem dpb provide a measure of natural giving-up densities (GUDs) and can be used to assess foraging costs and efficiencies, with greater stem dpb corresponding to lower costs and greater efficiencies. The results showed that stem dpb, bite rates, induced vigilance, and proportion of time spent in feeding differed between habitats. In particular, mountain nyala had greater stem dpb, higher bite rates, and spent a greater proportion of their time in feeding and less in induced vigilance in the grasslands. In addition, adult females had the highest bite rates, and the browse species Solanum marginatum had the greatest stem dpb. Generally, grasslands provide the mountain nyala with several advantages over the woodlands, including offering lower foraging costs, greater safety, and more time for foraging. The study advocates how behavioural indicators and natural GUDs are used to examine the habitat use of the endangered mountain nyala through applying non-invasive techniques. We conclude that the resulting measures are helpful for guiding conservation and management efforts and could be applicable to a number of endangered wildlife species including the mountain nyala.