The aim of the recently EU-funded MammoGrid project is, in the light of emerging Grid technology, to develop a European-wide database of mammograms that will be used to develop a set of important healthcare applications and investigate the potential of this Grid to support effective co-working between healthcare professionals throughout the EU. The MammoGrid consortium intends to use a Grid model to enable distributed computing that spans national borders. This Grid infrastructure will be used for deploying novel algorithms as software directly developed or enhanced within the project. Using the MammoGrid clinicians will be able to harness the use of massive amounts of medical image data to perform epidemiological studies, advanced image processing, radiographic education and ultimately, tele-diagnosis over communities of medical "virtual organisations". This is achieved through the use of Grid-compliant services [1] for managing (versions of) massively distributed files of mammograms, for handling the distributed execution of mammograms analysis software, for the development of Grid-aware algorithms and for the sharing of resources between multiple collaborating medical centres. All this is delivered via a novel software and hardware information infrastructure that, in addition guarantees the integrity and security of the medical data. The MammoGrid implementation is based on AliEn, a Grid framework developed by the ALICE Collaboration. AliEn provides a virtual file catalogue that allows transparent access to distributed data-sets and provides top to bottom implementation of a lightweight Grid applicable to cases when handling of a large number of files is required. This paper details the architecture that will be implemented by the MammoGrid project.