This article examines the basic economics of scholarly publications, especially technical journals, and applies this model to the consequences of low-cost electronic publication. The article discusses the demand for scholarly publication: dissemination of new information to students, other researchers and professional users outside the educational community; reputation development by scholars and research institutions; and the evaluation of research personnel by peers and superiors. The key supply feature of scholarly publication is that some uses are public goods, and others have strong economies of scale. Electronic publication reduces duplication and storage costs, but does not have much of an effect on fixed costs, and so is a minor technological change purely from the perspective of costs. However, electronic publication is a major change in two respects: it radically alters the relative costs of enhancements to straight textual material, and so may change the content of publications, and it dramatically reduces the cost of unauthorized duplication. The article focuses on the latter problem, and explores some of its possible consequences.