The Liber nouus de anima rationali (LNAR) was finished in Rome, in 1296. According to this date, the LNAR by Ramon Llull must be considered the first book in a series of works which are directed at revising and improving the traditional sciences, most of them of Aristotelian origin. He does so by applying his “new” and definitive approach, the Art, to the different traditional disciplines, e.g. astronomy, physics, metaphysics, etc. In the case of the LNAR, Llull focuses on the prolific psychological tradition, based on the Aristotelian book De anima. This text was very much discussed and commented upon during the thirteenth and fourteenth centuries. Formally, the work is structured according to ten principal parts which correspond to the artistic regulae/quaestiones given in the Tabula generalis (1294). With regard to the content of the LNAR, its anthropology partakes in the medieval dualistic vision of man which considers human beings as a central element within the hierarchy of the created world, which is composed, in turn, by spiritual and corporeal beings. Combining both natures, man is a kind of microcosm which is meant to provide the salvation for all corporeal beings, insofar as he acts as a link between them and their creator, God. This mission is realized by the proper agere of the rational soul: to produce knowledge by remembering, understanding and loving God. These activities, which are indispensable for a virtuous life, are carried out by the soul’s three faculties: memory, intellect and will, which work together with the substantial principia of the rational soul, that is bonitas, magnitudo, duratio, potestas, etc. Both parts of the soul, i.e. faculties and principles, have a correlative structure consisting of three elements (form, represented by the suffix “-iuum”; matter, represented by “-bile”; and action, expressed by the infinitive), which allow for dynamism and interrelation between the different parts, in a specular relationship with God and his Trinitarian nature. In order to establish the critical edition of the LNAR, we have fully collated 22 Latin manuscripts from several European libraries. In addition, we have taken into account the two printed editions of the Latin text as well as the Catalan text, transmitted by one manuscript, which was edited in 1950. Since a small group of Latin manuscripts transmits some obvious translation mistakes, we have been able to show that the Latin text is indeed a translation of the Catalan text. Due to this dependence of the Latin text on the Catalan, we have followed in general the lectures of these manuscripts which come closest to the Catalan, except for cases of important mistakes of translation, grammar or comprehension. The resulting Latin text is very close to the Catalan, both syntactically and lexically; yet, it stands as a text on its own with its own personality.