Hosting the Olympics has long been claimed to bring a financial return on investment. When this cannot clearly demonstrated, the medal surplus associated with host status is usually highlighted. We investigate the magnitude of having a home advantage at the recent Summer Olympic Games (1996–2021) separately for each organising country and by gender. Beside the host effect, the ex-ante and post impact on the medal surplus is also investigated in the same way. We analyse this effect using three models at the level of total medals, and medals won by men, and by women. Because of the number of zero observation in the response variable, we employ a zero-inflated negative binomial estimator. Our results cast some doubt on the claim of a host effect of the Olympics: when we control for socioeconomic variables in the models, the host effect fades away. Any host effect is clearer for medals won by men. Ex-ante and post effects on host countries are detectable in some cases but also fade with the inclusion of control variables. Based on our results, the medal surplus associated with the hosting of the Olympics is less evident than reported in literature or public discourse.