The focus of the paper is on the use of computer corpora in language research. The historical background is touched on, with special reference to work within the International Computer Archive of Modern English (ICAME). Developments in the use of corpora are surveyed. Issues taken up include the representativeness and structure of corpora. Special attention is paid to pitfalls in the use of corpora. Corpus compilers must provide adequate documentation on the texts. Corpus users must know the corpus in order to evaluate whether it is appropriate for their research problem and in order to evaluate the results of their studies.