Processes of naturalistic reading and writing are based on complex linguistic input, stretch-out over time, and rely on an integrated performance of multiple perceptual, cognitive, and motor processes. Hence, naturalistic reading and writing performance is nonstationary and exhibits fluctuations and transitions. However, instead of being just complications for the analysis of such data, they are also informative about cognitive change, fluency, and reading or writing skill. To use and quantify such dynamics, one needs appropriate statistics that capture these aspects. In this article I introduce Recurrence Quantification Analysis (RQA) as a tool to capture such dynamic structure. After a conceptual introduction of the analysis, I present a step-by-step tutorial on how to run RQA using R. Guidance is given with regard to common issues and best practices using this time-series analysis technique. Finally, I review previous results from studies applying RQA to reading and writing and summarize current hypotheses and interpretations of RQA measures in the context of reading and writing.