The world's most comprehensive digital biodiversity database has been developed, enabling access to over 177 million records. However, over two-thirds of the records are from just three countries: the US, Sweden and the UK, and focus on certain groups such as birds. Changes in policy and funding could help widen the scope of the data. In 2001 the Global Biodiversity Information Facility (GBIF) was formed to develop a portal to enable free and open access to primary biodiversity data. Primary data provides information on the location of animals or plants at a particular time and is useful for understanding the drivers of biodiversity loss, such as invasive species. The GBIF consists of 50 member countries and 40 member international organisations. Fifty-two of these have published datasets via GBIF and, as of July 2009, more than 177 million primary biodiversity records covering more than a million species have been published. However, the records are mainly from North America and Europe with 42 per cent coming from US, 15 per cent from Sweden and 13 per cent from the UK. Although Africa, Asia and Oceania (lands in the Pacific Ocean) have high levels of biodiversity, they are severely under-represented, contributing only 1.2 per cent, 1.6 per cent and 2.8 per cent respectively of the total GBIF records. A study of the data also found a bias in the different groups of animals and plants. Well-studied groups, such as birds, comprise 38 per cent of the GBIF records, although they only represent 0.6 per cent of the total number of species described by taxonomists worldwide. Insects comprise only 9.5 per cent, despite representing more than 60 per cent of total species described by taxonomists. The study suggests that a more comprehensive analysis is needed to categorise where information is missing. For example, research into ecosystem types, such as mangroves, and into taxonomic groups, such as the legume plant group, as well as geographic locations.