We apply the hedonic approach to the office property in Ankara and examine office rent determinants and their effects on rental variations. A hedonic regression model is utilized for the estimation of rent indices by using cross-sectional data including 251 office units in Ankara. Two different models were constructed in order to identify and evaluate the impact of location on the rental variations. The results identify the attributes (i.e., existence of a CPI escalation in the lease and vertical location of the office unit in the building) as proxies for the locational characteristics. Findings also suggest that locational characteristics have the greatest effect on the rental prices of the office units. The results of non-nested tests support the model incorporating locational attributes. Overall, the locational characteristics explain the spatial rent variations of the office property in Ankara to a large extent.