This chapter concerns the construction of a search engine to assist linguists, as well as other arts and humanities researchers working with text, to draw on the World Wide Web (web) as a source of linguistic information. The specially designed, standard corpora which underpin linguistic study are limited in the view of language use that they can provide, both in terms of size and of recentness, and the web is in principle freely available to provide supplementary and complementary data to fill the gaps. The first generation of web search tools, such as our ‘WebCorp’ system, has been dependent on commercial search engines to retrieve relevant text, a situation which has restricted the amount and type of information and the kinds of analysis and formatting which can be provided for the user. A second-generation web search tool, ‘WebCorpLSE’is being developed as a home-grown web search engine, tailored to bypass the obstacles presented by the non-language-geared and commercially-prioritised versions, and better serve the needs of researchers seeking evidence of particular language use on the web.