We assess the benefits of diversifying a portfolio of commercial real estate assets across gateway and non-gateway markets, a topic of significant relevance to institutional investors. Using simulation analysis and property-level data for the U.S., we compare various performance metrics for portfolios containing buildings in gateway markets only, both in gateway and non-gateway markets, and in non-gateway markets only, respectively. Our results suggest that the risk-adjusted performance is similar across types of markets. Gateway markets have higher appreciation and total returns, while non-gateway markets exhibit higher income returns even after accounting for capital expenditures. Downside risk appears to be slightly greater for gateway markets than for non-gateway markets; however, full drawdown and recovery lengths tend to be shorter for gateway markets. Our results further show evidence of momentum in appreciation returns, although no differences exist across types of markets. Income returns also appear to affect real estate pricing significantly, this effect being stronger for non-gateway than for gateway markets. By considering a large spectrum of performance metrics in a realistic investment setting, the results of the paper should provide investors with valuable information when allocating funds across gateway and non-gateway markets. The paper also provides important insights regarding how best to define gateway markets.