Because English is the lingua franca of world trade, the language of commerce, finance, and economics is characterized by an ever-increasing use of Anglicisms. Polysemic English loan words are particularly problematic in translation, as their meanings do not always match across donor and receptor languages. An Anglicism may, for example, convey a subset of the senses expressed by the same word in English and/or it may convey meanings typically expressed by a synonymous English word. It is no wonder that translator trainees often get into difficulty when having to decide whether and how to translate an English word with an established Anglicism in Italian. This tutorial presents a corpus-based teaching methodology that draws on the data-learning approach devised by Tim Johns and aims to equip translator trainees with a kit of analytical tools for better understanding Anglicisms in cross- and inter-linguistic professional communication so that they can produce accurate and effective translations. After briefly reviewing recent studies of Anglicisms in Italian, I outline the main features of the proposed educational methodology and illustrate how it has been applied to the analysis of the lemma business in the Language for Specific Purposes (LSP) translation classroom.