Abstract
Research on chronic disease arising from occupational exposure typically requires that events widely separated in time be linked through records created and maintained by more than one institution or agency. Too little attention is paid, in creating record forms, to the mechanics of record linkage and to the nature of the information needed for studies of the effect of the workplace on health. Record linkage also suffers from a highly decentralized federal statistical system and from changes in privacy legislation and in public attitudes toward personal privacy, changes motivated by invasions of privacy that have arisen from activities entirely unrelated to medical research. If medical research capabilities are to match the information needs of society a more adequate balance will have to be struck between those needs and personal privacy.