Real Estate has always been an investment class of its own, firstly, due to different risk/return characteristics than stocks and interest bearing assets and secondly, due to specific interdependencies to other assets in terms of return. These interdependencies are likely to be small, be it on a national level, i.e. between stocks or bonds and real estate investments, or on an international level, i.e. between real estate investments in different countries. Although low interdependencies between asset returns are the basis for gains by international diversification, real estate investments are still not a major factor in internationally diversified portfolios. This is largely due to the fact that real estate investments have some special characteristics, mainly high transaction- and information costs. These characteristics imply problems when integrating real estate into international diversified portfolios. A way of alleviating these problems is to use Real Estate Investment Companies as a substitute for and indirect form of real estate investments. In this paper we investigate the potential performance improvement by integrating Real Estate Investment Companies into internationally diversified portfolios consisting of conventional assets (stocks and bonds). We use an ex-post and an ex-ante approach. We utilize monthly return-series data on Real Estate Companies, as well as on common stocks and bonds, from France, Germany, Switzerland, the UK and the USA from 1980-1998. Our analysis uses the traditional Markowitz approach and, alternatively, a shortfall approach, due to the rather critical normality assumptions of the former. The analysis incorporates the effects of either hedged or unhedged exchange rate risk. The German and the US viewpoint leads to an examination of the dependency of the potentials for diversification on the investors home country. From a German investors point of view the integration of real estate investments has non-negligible effects on the performance of an internationally diversified portfolio, both in the ex-post and in the ex-ante case. These effects are mainly, independent of the utilized measures of risk, a large potential reduction of risk. The qualitative results for the US investor are similar to those in the German case. The quantitative effects however are less pronounced and are of less regular occurrence.