This work seeks to describe the lexicon of the beautiful and the ugly, in portuguese, using information from dictionaries to search a large corpus of literary texts. Specifically, the work makes use of the annotated corpora accessible by the AC/DC Project (Access to corpora/ Availability of corpora), in order to, from patterns that take into account lexical items in the field of beauty and ugliness, reveal what, in portuguese language, is considered beautiful and, also, what is considered ugly. Carried out in two different moments, the present work used a digital dictionary and a physical dictionary, in its first moment, in order to get the vocabulary of the portuguese language that circumscribes the semantic fields of beauty and ugliness and the Literateca corpus, in the second moment, which allowed the observation of the recurrent contexts for the lexical items of each field. Finally, in addition to the description and analysis of the data found, a cultural panorama is elaborated that tries to explain such occurences in the circumstances in which they appear.