The calculation of value-at-risk (VaR) has long been a problematic topic for real estate portfolios. The difficulties arise mainly from the lack of suitable data, the lack of transactions, the non-normality of returns and the special character of real estate. In addition, specific risks remain hidden in the real estate portfolios, so risk measurements based on standard methodologies do not exactly reflect the risks of a real estate portfolio. In this paper, different VaR models are examined for their suitability for estimating the risks of real estate portfolios. In the second step, the specific properties of real estate and the associated influencing factors on the risk are examined and approaches to measurement are discussed. The paper contributes to the literature on real estate risk management by discussing methods for estimating the risk of real estate portfolios and considering the characteristics of real estate investments. It provides approaches for more precise measurement of real estate risk.