All environmentally relevant information is computer stored for each organic chemical manufactured in quantities above 1.000 tons per year (whether it is considered as potentially toxic or not), known toxic chemicals (independently of their production figures) and certain classes of inorganic chemicals. The main data fields ("attributes" for each chemical) are: 1. substance identification and physicochemical data; 2. production and use data; 3. handling, transportation, and disposal; 4. environmental involvement (breakdown under biotic and abiotic conditions, average concentrations in environmental matrices); 5. biological effects and toxicity; 6. protection standards, administrative and regulatory data. Structure and substructure codification permits screening for substances with similar structures and coupling to GC-MS-computer analysis units with full or approximative structure outputs. A pilot project has been started permitting queries via direct and inverted file. It will include all available information on a limited number of compounds (some 5.000) from different chemical and use classes. The further software development foresees automatic indexing, automatic encoding of formatted data, automatic query formulation and search on a probabilistic basis for natural-language source data. Easy data flow is ensured by a networking between the central unit at the EC research Centre in Ispra (Italy) and specialised national data centres (ECDIN = European Chemicals Data and Information Network).