Are European Geographical Indications effective in the presence of a weak trademark mechanism and uneven application costs? I develop a multi-stage game of incomplete information to understand the impact of these forces on EU food quality standards. Under Decentralised Surveillance and pre-set charges, a social planner cannot reliably attribute trademarks to deserving producers and separate the market. While a Riley Outcome is unattainable, three equilibria survive the notion of sequential rationality. In showing that these outcomes are all suboptimal, I argue that EU regulators can improve the current application process by i) yearly reviewing the testing device and ii) replacing ex-ante with ex-post filing fees. Under its aegis, the WIPO could subsume both reformative proposals.