Robo-advisory is a new category in portfolio management and the investment management industry. Few studies have been done on how robo-advisors’ perform in the long run. The purpose of this research is to replicate and backtest the Swedish robo-advisors’ from 2010 to 2019 and analyse their performance. Data is collected from several assets that represent the actual robo-advisors’ underlying assets. The collected data is tested through a correlation test to ensure that it accurately represents the real robo-advisors’ portfolios and performance. The portfolios’ are recreated and then backtested through the use of the online software Portfolio Visualiser applying mean-variance optimisation and the Black-Litterman model to allocate the portfolio assets. The research successfully replicates the robo-advisors’ portfolios and finds that one robo-advisor, Lysa, outperforms the alternatives in most risk settings and continuously delivers one of the highest risk-adjusted returns. The main contributing factor to differences in performance is found to be the proportion of stocks and bonds in the portfolios.