The empirically documented Pareto wealth distribution at high wealth levels implies rather extreme wealth inequality. Is this inequality primarily due to differential talent, or is it due to luck? The answer to this question has profound political, social, and philosophical implications, as well as implications regarding market efficiency. We address this question theoretically and with a unique investment experiment with equal initial endowments and real out-of-pocket money. We show that the empirically observed Pareto distribution implies that luck, rather than differential investment talent, is the main force driving inequality at high wealth levels.