The current study aimed at investigating the authorial identity of Iranian academic writers, who came from three different fields of English, Biology, and Engineering, plus examining the influence of disciplinary conventions on their stance taking in research articles. The main objectives of this study were achieved by going through two main phases, viz. survey administration and corpus study. First, the authorial identity questionnaire was administered to 150 academic writers, 50 from each of the selected fields. Following that, an analysis of variance (ANOVA) was run to locate the difference between authorial identities of academic writers among these groups. Second, as a complementary phase to survey administration, NVivo was utilized to conduct the corpus study phase. In so doing, Hyland’s (2005) model of interaction in academic discourse was applied to analyze academic writers’ stance taking in a corpus comprising 90 articles from the three selected fields. Triangulating the findings, we concluded that academic writers in the field of English rely more on authority, self-representation, and personal projection, while those in the fields of biology and engineering try to take less stance markers and portray their findings more impersonally.