Most archival institutions now have large backlogs of unprocessed materials, and the burgeoning of digital media and digital information seems likely to exacerbate these backlogs if archivists continue to rely on traditional methods of description. In the era of digital abundance, new approaches to archival description will be required, and several possibilities have been suggested in recent years. This paper examines four of them: minimal processing; reuse of metadata supplied by records creators; user contributions; and automated methods of capture of descriptive information. It considers how far each of these approaches is practical and which of them may offer the most promise in the digital world. The paper concludes that, in the future, automated capture will probably meet many of our needs for the supply of documentation and will expedite work that would otherwise be impossibly time-consuming. However, it will almost certainly have to be used in conjunction with other methods that rely on human input.