In very broad strokes, as indicated by the International Union for the Conservation of Nature and Natural Resources (IUCN) in 1980, biology can be thought of at three levels of organization: molecular/genetic, species and ecosystem.The raw data of the molecular level are nearly all digital, as are many of those at the ecosystem level.However, the raw data of the species level (where they are found, the physiology, morphology, etc.) are almost all entirely analogue and descriptive.However, developments in informatics at each of these levels can be of service to the others.The Global Biodiversity Information Facility (GBIF) was established to enable the digital capture and dissemination of data related to natural history specimens (including those in culture and other living collections), of which there are an estimated 1.5 billion in at least 6000 collections worldwide.Another of GBI F 's tasks is to generate an electronic catalogue of names of known organisms, which is the element required to enable data mining across all three levels in a single quer y.GBI F 's work at the species and specimen levels of biological organization can be thought of as unifying the biological information domain.In addition, it provides worldwide coordination among the many ongoing digitization projects, standards development and networking efforts within biodiversity informatics. WHAT IS GBIF?The Global Biodiversity Information Facility has a mission to make the world's species' biodiversity data freely and universally available via the Internet.It is a megascience facility -in part because the GBIF concept was developed by a working group formed by the Mega Science Forum (now the Global Science Forum) of the Organisation for Economic Cooperation and Development.More importantly, it is megascience because it is a worldwide endeavour that is challenging in the several areas of information science, technology and sociology as well as biology.GBI F 's efforts are focused on primary scientific biodiversity data at the specimen and species levels because these data, unlike most molecular/genetic and much ecological data,