This study investigated the linguistic features of abstracts across five scientific and technical disciplines: biochemistry, civil engineering, computer and information sciences, electronics engineering, and mechanical engineering. Abstracts play a crucial role in academic papers by summarizing key findings, methodologies, and implications. Through the lens of English for Specific Purposes (ESP) and genre analysis, this study aimed to determine whether linguistic features vary across disciplines and classify these disciplines based on the similarity of their linguistic features. The corpus consisted of 5,000 abstracts, with 1,000 from each discipline, sourced from the open-access journal PLOS ONE. Using Biber's [1] multidimensional analysis framework, this study examined 63 of 67 linguistic features, including passive voice, conjunctions, amplifiers, and discourse markers. Statistical analysis, including correlation and cluster analyses, revealed that the disciplines can be broadly divided into two groups: biochemistry and computer & information sciences, and a second group including mechanical engineering, civil engineering, and electronics engineering. These findings suggest that while some linguistic features are shared across disciplines, others vary substantially. For example, biochemistry has a higher frequency of passive constructions and large noun phrases, whereas computer and information sciences frequently use first-person pronouns and amplifiers. These insights are valuable for ESP instruction as they highlight the need for discipline-specific writing guidance in higher education. Educators can use this information to develop effective writing instruction tailored to the linguistic norms of each field. Future research could expand on these findings by exploring additional rhetorical elements and examining the impact of linguistic features on reader comprehension.