The term neuroinformatics is often used to refer to the application of information technology (IT) to the "brain sciences." Almost all of the brain sciences have become considerably more complex, and recording and managing the results from experiments entails the use of ever-more complex and larger databases and analysis tools. However, inspite of our knowledge of the brain and brain function rapidly expanding, our ability to make use of this information is not increases at the same rate. This among other considerations have led to the creation of nEUroIT.net (http://www.neuro-it.net), a thematic network in the area of NeuroIT. NeuroIT has here been defined loosely as "neuroscience for IT," whereas in "neuroinformatics" the emphasis has more been on "IT for neuroscience." The distinction, therefore, is not so much in the field of study, or in the techniques used, but rather in the long-term objective of the research. nEUro-IT.net is funded by the Future and Emerging Technology (FET) arm of the Information Society Technology (IST). One of its most important activities is the creation of a roadmap. There are a number of reasons for a roadmap. A roadmap tries to develop a vision for where the field will be in the next decade. It tries to identify problems that affect the field as a whole and how they can be solved. It serves as a reference for the state-of-the-art research in various fields for researchers and for decision makers. This is perhaps even more important for NeuroIT than for other fields, since it is highly interdisciplinary: a computer scientist cannot be expected to know about the latest developments in primate vision, and yet these developments may provide crucial inspiration for new approaches in computer vision.