Even though Brazilian Portuguese (BP) has replaced its third person accusative clitic pronouns (ClA3) for strong pronouns, they have not completely disappeared: although they are not acquired during the first years of BP native speakers’ linguistic development, they are a topic taught in Brazil classrooms. However, they seem to behave differently from the other clitics (which occur in proclisis to the main verb). Nunes (2015) suggests that they have been reanalyzed as object agreement marks, given their tendency to be placed in post-verbal position of verb forms that can independently carry subject agreement marks, to occur in proclisis with verbs whose subject agreement mark is occupied, and not to occur next to verbs in the past participle or in the gerund, since they do not have a position for agreement. In order to control some linguistic and sociolinguistic variables, this study had the objective to gather experimental data regarding the idiosyncrasies the ClA3 have using a sentence rewriting task including the ClA3 “o” and the clitic “me”. There were differences between the use of these two pronouns, having the clitic “me” followed the general tendency of BP, being placed, almost exclusively, in proclisis to the verb it is an argument of. Enclisis, on the other hand, was more productive with the ClA3, namely with non-inflected infinitive verbs. However, proclisis to the main verb was also a widely used option by the participants, including in contexts Nunes (2015) did not predict it to occur. It is concluded that, at first sight, Nunes’s (2015) hypothesis was not backed by the data obtained here. Nevertheless, the difference in the speakers’ behavior regarding the two pronouns used in this study cannot be disregarded, what suggests they are likely to have a different nature, although they do not seem to be object agreement marks.