Both the treaties of the European Union (EU) and the secondary legislation from EU institutions are currently issued in twenty-three different language versions. Discrepancies between these versions both jeopardize their equal authenticity, and make a uniform interpretation and application of EU law in all EU Member States more difficult. Yet, discrepancies between language versions have come up in the case law of the European Court of Justice (ECJ). Using a statistical analysis of fifty years of the ECJ's case law, all relevant judgments between 1960 and 2010, this article sheds further light on which interpretive method the ECJ is more likely to use in the event of linguistic discrepancies as an argument to justify its interpretation of EU law. It first discusses the discrepancies between the language versions that impede the multilingual interpretation of EU law. The article then makes use of the so-called ‘legal reasoning’ methodology to analyze the ECJ's case law, and considers the ECJ's use of the teleological approach and the literal approach to discrepancies.