Automated, frequency-driven approaches to identifying commonly used word combinations have become an important aspect of academic discourse analysis and English for academic purposes (EAP) teaching during the last 10 years. Referred to as clusters, chunks, or bundles, these sequences are certainly formulaic, but in the sense that they are simply extended collocations that appear more frequently than expected by chance, helping to shape meanings in specific contexts and contributing to our sense of coherence in a text. More recently, work has extended to concgrams, or noncontiguous word groupings where there is lexical and positional variation. Together, these lexical patterns are pervasive in academic language use and a key component of fluent linguistic production, marking out novice and expert use in a range of genres. This article discusses the emerging research which demonstrates the importance of formulaic language in both academic speech and writing and the extent to which it varies in frequency, form, and function by mode, discipline, and genre.