With the beginnings of a worldwide burgeoning development of matched firm-employee data, it is worthwhile to examine the possibilities for using these data. This essay discusses a variety of areas in which some progress has been made and presents ideas for future research in a number of others, including the study of labor demand, search and unemployment, wage determination and time use. It concludes that such data could be as important for labor economics, and for generating new knowledge about labor markets, as have been longitudinal household datasets, but with existing restrictions on access this kind of success will be difficult to achieve.