This chapter presents a brief outline of the relationship between data and language documentation. Three kinds of resources have long been given a special place in descriptive linguistics: texts, dictionaries and grammars. It has become standard practice for linguists documenting under-resourced languages to consider ways in which their work can result in outputs not only for use in academic spheres, but also community ones. The chapter explores the notion of an underlying data structure and introduces general aspects of the problem of encoding that structure in machine-readable format. It covers specific issues relating to the encoding of language data on a computer. Most of the documentary objects requiring metadata can be arranged in a hierarchy from more general to more specific using the categories project, corpus, session and resource. The chapter concludes with a discussion of linguist's responsibilities for navigating the relationship between their data and new technologies.