The DIA-DB is a web server for the prediction of diabetes drugs that uses two different and complementary approaches: (a) comparison by shape similarity against a curated database of approved antidiabetic drugs and experimental small molecules and (b) inverse virtual screening of the input molecules chosen by the users against a set of therapeutic protein targets identified as key elements in diabetes. As a proof of concept DIA-DB was successfully applied in an integral workflow for the identification of the antidiabetic chemical profile in a complex crude plant extract. To this end, we conducted the extraction and LC-MS based chemical profile analysis of Sclerocarya birrea and subsequently utilized this data as input for our server. The server is open to all users, registration is not necessary, and a detailed report with the results of the prediction is sent to the user by email once calculations are completed. This is a novel public domain database and web server specific for diabetes drugs and can be accessed online through http://bio-hpc.eu/software/dia-db/.