Knowledge of species’ habitat requirements is important for planning conservation programs for the target species. The accessibility of habitat requirements of a species is know as habitat suitability. Wild goat ( Capra aegagrus ) is a vulnerable species which is widely distributed in Iran , but the ecology of this species is not sufficiently studied. The purpose of this study was modelling seasonal habitat suitability and habitat associations of wild goat in Kolah-Qazi National Park , one of its typic habitats in cental Asia . A two-step approach was employed for the modelling process. Presence only data was used in the first step to determine unsuitable sites employing Maximum Entropy approach. Absence points were then randomly selected from the determined usuitable sites by the model. Present points had been seasonally collected across the study area in a previous study (Shahgholian, 2012). Logistic Regression approach was used in the second step to model wild got habitat suitability using both the presence and absence points. Elevation, aspect, distance to water trough, distance to settlement and distance to guard patrol road were recognised as the most important variables determining habitat suitability of wild goat. According to the produced seasonal habitat suitability maps, the extent of suitable habitats provided by the logistic regression model was more than the MaxEnt model. The extent of suitable habitats was maximum in winter and was least in summer. The accuracy of both MaxEnt and LR model were good, but of the LR model was generally better. The obtained results have implications for conservation of wild goat across its distribution range. Key word: Habitat suitability model, Capra aegagrus, Two-step approach, Logistic regression, Maxent, Kolah-Qazi National Park