The re-use of existing and official data has a very long and largely honourable history in education and social science. The principal change in the 60 years since the first issue of the "British Journal of Educational Studies" has been the increasing range, availability and quality of existing numeric datasets. New and valuable fields of endeavour have formed around specific applications of these datasets, and some promising analytical techniques have been devised to deal with them. At the same time, the opportunities provided by these datasets have thrown up fascinating methodological and other challenges. This paper presents a brief summary of all of these developments, with illustrative examples. The paper then considers two fields of endeavour that have been particularly valuable in education studies--the political arithmetic tradition of origins and destinations, and school effectiveness and improvement. Both have used similar datasets but for different purposes and so reaching rather different conclusions, with considerably different take-up by policy-makers. The paper ends by envisaging some of the ways in which these and other fields using secondary data may develop in the near future. (Contains 1 note and 1 table.)