this study purported to compare and contrast the use of self-mention and evidentials as two mtadiscourse features in opinion columns of persian and english newspapers. the theoretical basis of this study is the idea that metadiscourse features vary across cultural boundaries. for this purpose, 150 persian and 150 english opinion columns were collected based on three factors of topic, audience and nativity of the writer. with regard to evidential features, it was found that inferential and reportative evidentials were the main kinds of evidentials found in both persian and english opinion columns. reportative evidentials were divided into 7 subcategories based on four opposing characteristics that were found among them and including inferential evidentials there were 8 evidential categories in the end. reportative evidentials either were personal or institutional, generic or specific, identified or unidentified, authoritative or non-authoritative. the evidential sub-categories that were formed considering these characteristics were as follows: personal unidentified generic authoritative sources (puga), personal identified specific authoritative sources (pisa), impersonal non-institutional unidentified generic authoritative sources (inuga), impersonal non-institutional unidentified specific authoritative sources (inusa), personal unidentified non-authoritative sources (puna), institutional unidentified generic authoritative sources (iuga), institutional identified specific authoritative sources (iisa). it was found that persian and english opinion columns used pisa and iisa to the same degree and chi-square tests conducted with regard to these evidentials showed no significant difference. persian opinion columns showed a significantly higher frequency of unidentified authoritative sources of information that is puga, inuga, inusa evidentials. this observation was interpreted as a sign that persian columnists opted for an authoritative style of writing. inferential evidentiality was the only evidential feature that was found to be significantly more frequent in english opinion columns. higher frequency of inferential evidentials in english opinion columns was interpreted as an indication that english columnists preferred a subjective mode of writing. findings about self-mention features and the discourse functions accomplished through them further supported this interpretation. the total number of self-mention features in english opinion columns was more than twice the number of these features in persian ones. the types of discourse functions accomplished through using self-mention features were also significantly different across the two corpora. while persian columnists mostly used self-mention features for organizing the text and introducing discourse steps, their english counterparts used such features for expressing personal beliefs, illocution function, emotional appeal to the reader and clarifying their intention. along with the tune of subjectivity that these self-mention features and their discourse functions bring in the text, such features can also be viewed as an indicator of degree of self- assertion on the part of the writer.