Among the various theoretical-clinical developments put forward by Balint, which are sometimes extensions and deeper explorations of Ferenczi's theorizations, but at other times innovations put in place by the great researcher and solid clinician that he was, I have chosen the question of regression and its handling in analysis, a question that condenses several others, two of which are essential : the three psychic spaces and the basic fault. If we can clearly state that Balint learned a great deal from Ferenczi, both from his discoveries and successes as well as from his errors and failures, it is around regression that this question is most pertinent. We will see here, essentially, that Balint cannot be considered as someone who simply continued the path initiated by his master. The question of regression and its handling by the pupil is a real masterstroke !