Research into the recruitment processes for European elections has tended to focus on either the political resources of candidates, providing 'supply-side' explanations, or the peculiarities of the multi-level political context, determining the 'structure of opportunities' for these recruitments. The analysis of party rules and of the attitudes of gatekeepers, producing the 'demand' for candidates has often been overlooked or unsystematically studied. National parties are often assumed to be unitary actors which unilaterally control ballot access for European elections, although the actual processes remain empirically unexplored. Based on a new dataset recording the candidate selection methods used in the various political parties of the 28 member states in view of the 2014 European Parliament elections, this research examines how EP candidates are selected. Formal rules prevailing in the parties are collected through the organisations' internal documents (statutes, resolutions) and an expert survey. Informal practices are established through MEPs' interviews. In order to classify the methods thus unveiled, the paper relies on an analytical framework distinguishing the eligibility criteria, the selectorate and the decentralisation dimensions of candidate selection for EP elections, taking into account the multi-stages nature of the processes. Findings allow in particular for cross-party and cross-national comparisons.