We test whether the inclusion of green asset ETFs in portfolios yields better outcomes for investors. We use a mean-variance optimization framework to construct optimal portfolios with and without green assets and compare their performance over different time horizons and market conditions. Our results show that the portfolios that combine green assets with other broader market indices generate higher returns and Sharpe ratios compared to benchmark portfolios. These results survive the incorporation of transaction costs and removal of the COVID-19 period from the sample. Further, we find that the probability of outperforming the benchmark is much higher for long-horizon investors in green asset portfolios.