Frequency lists are useful in their own right for assisting a linguist, lexicographer, language teacher, or learner analyse or exploit a corpus. When employed comparatively through the keywords approach, significant changes in the relative ordering of words can flag points of interest. This conceptually simple approach of comparing one frequency list against another has been very widely exploited in corpus linguistics to help answer a vast number of research questions. In this chapter, we describe the method step-by-step to produce a keywords list, and then highlight two representative studies to illustrate the usefulness of the method. In our critical assessment of the keywords method, we highlight issues related to corpus design and comparability, the application of statistics, and clusters and n-grams to improve the method. We also describe important software tools and other resources, as well as providing further reading.