Data sharing is something of a buzzword in current scientific research, and the neurosciences have not been impervious to its influence. In particular, the idea of sharing neuroimaging data has gained significant support, and several efforts are already in place to ensure that this commendable goal is achieved. But although everyone agrees that sharing imaging data is a great idea, there is less of a consensus about the best way to implement such a database. There are many questions that need to be answered before we advance at full speed on this front, questions that range from methodological issues to strictly ethical problems. What type of data do we want to share? How will we monitor quality control? What kind of data-mining capabilities will we include in the database and who is going to develop them? How are we going to protect intellectual property and the rights of the authors? How are we going to protect the identity of subjects from whom the images were obtained? Any serious attempt to develop an imaging database must take into consideration these and many other questions.

Fortunately, neuroscientists are well aware of these problems and have begun to tackle them. In this issue, we feature five articles that address these questions, pose new challenges, and present some plausible solutions. As the endeavours of the imaging community are breaking new ground, it is in the interests of other neuroscientists to understand and evaluate these efforts. All areas of scientific research are moving towards the idea of developing the ultimate 'big picture' by integrating data across all levels of analysis. Owing to its pioneering nature, we should not be surprised if the thinking of the imaging community traces a path that other neuroscientists will have to follow.