Turtles are a globally threatened group of reptiles. Zoo populations may contribute to the conservation of species, including turtles, but collection composition may not align with conservation needs. We combined data from the Zoological Information Management System (ZIMS), EDGE of Existence, the IUCN Red List and the Reptile Database to investigate zoo turtle holdings on global and regional scales. Globally, zoo collections were representative of turtle diversity, regional species distributions and threat statuses, indicating no bias towards threatened species and no taxonomic or distribution blind spots. Species kept in zoos had significantly lower EDGE scores than those not represented, and threatened species were no more likely to have been bred in the year prior to data collection (before March 2022) or have non-viable populations, but were more likely to have a larger population size. Although Africa, Asia and South America have the smallest turtle holdings in terms of species, allowing for regional capacity, these regions hold more, while Europe holds fewer than expected turtle species – North American and Asian holdings do not differ from expected. African, Asian, North and South American regions significantly bias their collections towards native species. We found evidence for significant increases in turtle populations at the genus level following the EAZA Shellshock campaign in Europe. ZIMS data are limited by taxonomy, membership and accuracy of records but provide the best window into patterns of zoo turtle collections. While holding a species in a zoo does not equate to conservation value, based on these data, we recommend that conservation prioritisation exercises are developed for all turtle species, holding institutions or regional taxonomic advisory address population viability and support for institutions working with significant turtle populations in captivity to join ZIMS is provided. Keywords: chelonia, ex-situ, zoos, conservation, ZIMS