The present study treats translation errors which University students are capable of recognizing and correcting. A translated text in Greek, containing errors made on purpose, was proposed to students, who were asked to intervene. Some of those errors were successfully identified and rectified, while others were not. In some cases, the students intervened by rectifying translation units that were absolutely correct. The major hypothesis is whether this kind of exercise helps students realize dissymmetrical morphosyntactic and lexical uses and to what extent. We argue that the interest of such type of exercise is obvious, given that students act like readers, thus feeling more comfortable during the task assigned and, therefore, identify errors easier. As far as our results are concerned, it appears that our hypothesis is generally verified, although subject to some constraints.