Today, hardly any political debates occur without accusations of someone making a populist argument. However, it is not yet clear whether populist rhetoric is empirically really as common as one might think and, if so, what its effects are on parliamentary debates and legislative behaviour. To answer these questions, I quantitatively analyse more than 300,000 speeches given in the European Parliament between 1999 and 2017 by combining a novel supervised machine-learning approach based on a crowdsourced sample of speeches with word embeddings, dictionaries and unsupervised topic models. In addition, I incorporate longitudinal public opinion data from 28 European countries to show that elites use populist rhetoric strategically and that the presence of populist rhetoric in parliament depends on factors such as issue salience, polarisation and party positions. The contribution of this paper is twofold: Methodologically, it develops a novel, quantitative measure for populist rhetoric in parliamentary debates. Substantially, it thoroughly investigates the effects of populist rhetoric on parliamentary debates as well as voters and thus how populism transforms European Politics.