The research project GLOWA-Danube, financed by the German Federal Government, is investigating long-term changes in the water cycle of the upper Danube river basin (77,000 km(2)) in light of global climatic change. Its aim is to build a fully integrated decision-support tool "DANUBIA" that combines the competence of 11 different research institutes in domains covering all major aspects governing the water cycle-from the formation of clouds, to groundwater flow patterns, to the behaviour of the water consumer. Both the influence of natural changes in the ecosystem, such as climate change, and changes in human behaviour, such as changes in land use or water consumption, are considered. DANUBIA is comprised of 15 individual disciplinary models that are connected via customized interfaces that facilitate network-based parallel calculations. The strictly object-oriented DANUBIA architecture was developed using the graphical notation tool UML (Unified Modeling Language) and has been implemented in Java code. All models use the same spatial discretisation for the exchange of data (1 x 1 km grid cells) but are using different time steps. The representation of a vast number of relevant physical and social processes that occur at different spatial and temporal scales is a very demanding task. Newly developed up- and downscaling procedures [Rojanschi, V., 2001. Effects of upscaling for a finite-difference flow model. Master's Thesis, Institut fur Wasserbau, Universitat Stuttgart, Stuttgart, Germany] and a sophisticated time controller developed by the computer sciences group [Hennicker, R., Barth, M., Kraus, A., Ludwig, M., 2002. DANUBIA: A Web-based modelling and decision support system for integrative global change research in the upper Danube basin. In: GSF (Ed.), GLOWA, German Program on Global Change in the Hydrological Cycle Status Report 2002. GSF, Munich, pp. 35-38; Kraus, A., Ludwig, M., 2003. GLOWA-Danube Papers Technical Release No. 002 (Danubia Framework), Software-Release No.: 0.9.2, Documentation Version: 0.10, Release Date: 27 March 2003] are required to solve the emerging problems. After a first successful public demonstration of the DANUBIA package (nine models) in May 2002 [Mauser, W., Stolz, R., Colgan, A., 2002. GLOWA-Danube: integrative techniques, scenarios and strategies regarding global change of the water cycle. In: GSF (Ed.), GLOWA, German Program on Global Change in the Hydrological Cycle (Phase I, 2000-2003) Status Report 2002. GSF, Munich, pp. 31-34], the research consortium is now preparing a first validation run of DANUBIA for the period 1995-1999 with all 15 models. After successful completion of the validation, a scenario run based on IPCC climate scenarios [IPCC, 2001. Climate Change 2001: Synthesis Report. In: Watson, R.T., Core Writing Team (Eds.), A Contribution of Working Groups I, II, and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK and New York, NY, USA, 398pp] for a five year period between 2025 and 2040 will follow at the end of 2003. The research group "Groundwater and Water Resources Management" at the Institute of Hydraulic Engineering, Universitat Stuttgart, is contributing both a three-dimensional groundwater flow model of the catchment and an agent-based model for simulating water supply and distribution. This paper gives a general overview of the GLOWA-Danube project and describes the groundwater modeling segment. Nickel et al. deal with the water supply model in a second contribution to this special issue. A three-dimensional numerical groundwater flow model consisting of four main layers has been developed and is in a continual state of refinement (MODFLOW, [McDonald, M.G., Harbaugh, AW., 1988. A modular three-dimensional finite-difference ground-water flow model: US Geological Survey Techniques of Water-Resources Investigations, Washington, USA (book 6, Chapter A1)]). One main research focus has been on the investigation of upscaling techniques to meet the requirement of a fixed 1 x 1 km cell size. This cell size is compulsory for all models in DANUBIA in order to facilitate a one to one parameter exchange. In a second stage, a transport model (nitrogen) will be added (MT3D): [Zheng, C., Hathaway, D-L., 1991. MT3D: a new modular three-dimensional transport model and its application in predicting the persistence and transport of dissolved compounds from a gasoline spill, with implications for remediation. Association of Ground Water Scientists and Engineers Annual Meeting on Innovative Ground Water Technologies for the '90s, National Ground Water Association, Westerville, Ohio, USA. Ground Water 29 (5)]. (c) 2005 Elsevier Ltd. All rights reserved.