The discussion section of a research article is the last section in which non-native researchers, especially in humanities and social science, face problems in determining its pattern and structure. In addition, there are fewer studies on this subject of discussion section in research articles in the domain of humanities and social sciences. This study attempted to understand the pattern and structures of the discussion section in research articles for easy and convenient writing of this section. The population consisted of 30 research articles published in the Scopus database and belonged to Quartiles 1-3 between 2004 and 2018. All research articles were in humanities and social sciences and included three field groups: language, linguistics, and language teaching; business, management, accounting, economics, marketing, and finance; and other fields of humanities and social sciences. The data were analyzed by grammatical theories, text analysis by Ure (1989), and the move structure model adopted from J. M. Swales (1990, 2004), Kanoksilapatham (2005), and Peacock (2002). This research found that the text pattern analyzed by Ure (1989) was "Specialized / Written text / Monologue / Discussion / Exploring + Recommending; while the results of move structure revealed that there were seven moves that occurred in the discussion section. They were Move1: Preparing for the presentation of results, Move2: Reporting results, Move3: Stating comments on the results, Move4: Stating limitations of the study, Move5: Recommendation, Move6: Implications, and Move7: Claim. Two moves contained steps: Move1 and Move3. Only Move2 occurred obligatorily while the others were optional occurring. [Note: The page range (217-231) shown on the PDF is incorrect. The correct page range is 217-232.]