The aim of this paper is to present the methodology used to compile a business corpus and identify its terminology in order to create a glossary to be used in translator training. First, we briefly review the literature on corpus building and exploitation for terminological purposes. Then, we present the methodology itself and a series of activities aimed at acquiring business knowledge. The results show that both techniques used to exploit the corpus (term detection and extraction) were useful, even though they do not fit the specific needs of this study. On the other hand, the exercises can be complemented with other exercises and modified according to the teaching context.