Anthropologists often have interesting and valuable data that remains ‘hidden’ because it does not fit easily into conventional academic publishing formats. This article suggests that it is worthwhile to make use of this hidden data for the benefit of other researchers and the study communities. To illustrate, the article describes initial efforts to create an online database of traditional weather prediction indicators derived from observations of the ecosystem. The database was started with descriptions of more than a thousand prediction indicators used in Northeast Brazil, which were collected as part of a survey of farmers and ‘rain prophets’. It is argued that such a database is important not only as part of the anthropological record, but also for the preservation of cultural heritage, and as a baseline for studies of environmental change. Some of the theoretical, practical, and ethical issues that have emerged in developing the database include: determining how much contextual information to include, obtaining translations, recruiting contributors, and properly acknowledging intellectual property. While there seems to be a great deal of enthusiasm for the idea from various sectors within and outside of academia, difficulties in securing funding for this interdisciplinary project and establishing a group of collaborators have so far presented significant obstacles.