Previous research on the home advantage in the history of the Olympic Games has found initial evidence that host nations have won more medals than non-hosts. In this paper, we argue that these findings are a myth of sports history, providing poor estimates of the home advantage in the Olympics. We argue that selection bias accounts for the findings in previous work, which uses an empirical strategy of comparing host nations to all non-hosts and to historical performances of host countries with much smaller delegations. When we correct for this bias the evidence in favour of a hosting advantage disappears. Additionally, we argue that the existing literature has largely neglected the rules about athlete qualification for host countries. To the extent that a small home advantage does exist, it is almost entirely driven by increased participation rates.