Diversification in real estate portfolio management can be mainly differentiated in geographic or property-type based strategies. Within geographic diversification, risk reduction can be achieved through different strategies. Diversification benefits through international diversification have been shown by several researchers. Many authors also suggest that regional diversification within one country would improve the risk-return-profiles of real estate portfolios. Lately, research on geographic diversification has focused on economic factors. Several studies have demonstrated that diversifying over regions that differ in their economic conditions (e.g. industry dominance, employment growth etc.), may lead to superior results compared to traditional naive geographic diversification. However, only few studies have examined the benefits of intracity diversification. This is surprising since investing in real estate always bears a trade-off between the benefits of wide geographic diversification and the costs of information and management of those investments. Investment managers often do, therefore, concentrate on few locations because of the high costs and efforts broad investment approaches imply. Due to the special characteristics of real estate properties (heterogeneity, fixed location), however, real estate markets are local markets and assets in local submarkets of a city may behave differently than the city in which they exist. Therefore, research on intracity diversification might shed more light on the real estate portfolio management process and may lead to higher portfolio efficiency by reducing the number of cities required to achieve diversification benefits. This study examines such benefits obtainable through intracity diversification. We employ a unique, twenty-six year (nine year) total return data set for London (Paris) and use quadratic programming to construct a diversified investment portfolio by optimally weighting the intrametropolitan submarkets. The results show that reductions in portfolio risk, represented by standard deviation, are obtainable through intracity geographic diversification at a given rate of return. Therefore, portfolio managers should not see cities as homogeneous markets and should not solely rely on aggregated data, as the latter does not reflect differences in local submarkets.