While members of the European Parliament are elected in national constituencies, their votes are determined by the aggregation of MEPs in multinational party groups. The uncoordinated aggregation of national party programmes in multinational EP party groups challenges theories of representation based on national parties and parliaments. This article provides a theoretical means of understanding representation by linking the aggregation of dozens of national party programmes in different EP party groups to the aggregation of groups to produce the parliamentary majority needed to enact policies. Drawing on an original data source of national party programmes, the EU Profiler, the article shows that the EP majorities created by aggregating MEP votes in party groups are best explained by cartel theories. These give priority to strengthening the EP’s collective capacity to enact policies rather than voting in accord with the programmes they were nationally elected to represent.