At the MPI for Psycholinguistics we launched a new web site (www.mpi.nl) on April 20, 2009. The web site is created in the Content Management System Plone. Our Plone CMS communicates with three data bases: 1. The MPI people data base: The interaction between the data base and the CMS is through a python script which allows us to create standard person pages for each MPI member. Person pages have a default person home set-up, a standard sub folder structure and some extra functionality to create user specific content. 2. The MPDL PubMan data base. Every night our CMS retrieves the metadata from the PubMan data base and stores these in the CMS using a python script. Researchers can specify a set of criteria to select the publications they wish to show on the publication folder of their person pages. The display can either be set to sort the publications by year, or by publication type or a combination of the two. A special (and hidden) 'local tag' element in the PubMan metadata set has been created to either create a collection for special publications or to exclude publications to be presented on the web page. In addition, research group and projects can create publications list based on the research group or project name. The MPDL metadata accomodates persistent links to the full text which can be viewed directly from the website when set at Public (in PubMan). In case the publication is set at Private a request form will appear with which users can send a request to our Library. The MPDL metadata also accomdates persistent links to the MPI-psycholinguistics research data archive. This allows the users to store supplementary material (primary data) in this archive and to link it to the PubMan data base. On the web site it will show automatically if the publiaction has supplementary material. 3. The MPI research data archive At our institute we house a primary research data archive. From the web site links can be created to these data, which will provide an alternative (and more public friendly) manner of showing the data. To realize the communication between the CMS and the three data bases has been quite a technical and organisational challenge, involving a lot of commitment from the MPDL team members and from the institute staff involved in the development of the web site (library, techical support). We feel that the functionality we have created is complex, but very nice and -easy-to-be-used- by other MPI institutes with a Plone based web site.