This paper reports a corpus investigation of the Methods sections of research-reporting articles in academic journals. In published pedagogic materials, Swales and Feak [Swales, J. M., & Feak, C. (1994). Academic writing for graduate students. Ann Arbor: University of Michigan Press; Swales, J. M., & Feak, C. (2000). English in today's research world. Ann Arbor: University of Michigan Press.], while not offering a generic structure, discuss the tendencies for Methods sections reporting research in the social sciences to be "slow" (or "extended"), and those in the physical sciences, such as medicine and engineering, to be "fast" (or "compressed")--the metaphors of speed or density relating to the degree of elaboration employed in describing and justifying the research design and process. The aim of this study is to examine the differences between fast and slow tendencies in Methods sections in terms of their internal, cognitive discourse organization. Two small corpora, each consisting of thirty Methods sections (one for each of the two groups of subjects), are analyzed in two ways. First the corpora are rater-analyzed for their use of the organizational features of a "cognitive genre" model for textual structures (see Bruce, I. J. (2005). Syllabus design for general EAP courses: a cognitive approach. Journal of English for Academic Purposes, 4(3), 239-256.) and secondly by the use of corpus software for linguistic features that characterize the model. The findings of the study suggest that "fast" Methods sections that report research in the physical sciences generally employ a "means-focused" discourse structure, and "slow" Methods sections in social science reports tend to employ a combination of "chronological" and "non-sequential" descriptive structures. The study concludes that learner writers may benefit from access to the types of general, "procedural" knowledge that these discoursal structures employ.