The TRY initiative (www.try-db.org) is a network of vegetation scientists providing curated plant trait data for the scientific community. The TRY Database currently contains about 7 million trait records for nearly 3000 different traits. The flexible database structure can hold any number of traits and a generic program can import any kind of data without requiring a template. About 10 million trait records for about 100 requests are released on a monthly basis. This is organized via the TRY Data Portal, which facilitates data contribution, exploration and customized requests. The Dataset Custodian Centre allows managing the status of contributed datasets and monitoring the use of these data from requests to scientific publications. The Request PI Centre allows managing and monitoring requests. Both centres facilitate direct contact of data contributors and users. In addition to the TRY Database we have established a file archive, which facilitates publication and DOIs for else unpublished plant trait datasets. The TRY Data Portal has evolved toward a long-term scientific data infrastructure, which combines the advantages of easy access to curated plant trait data almost ready for analyses, with direct contact of data providers and users, the opportunity for data providers to publish individual datasets and track the use of their data. This presentation will introduce details of the TRY database system.