Psychiatric neuroimaging has become a standard approach in the investigation of mental disorders. Many results have been reported,but a lot of them not been replicated. Some of this is due to underpowered sample sizes, some to statistical inappropriate approaches, and some just may be true negative findings. Currently, it is unclear, which results will stay or emerge that are reliable and valid. A recent move, similar as in genetics, has been made to big data approaches in either large, longitudinal cohorts or by collecting imaging data across studies and continents, combined with ecological momentary (mobile) assessment. This does create big data that have to be analyzed properly. Although new methods have been developed and rediscovered the community has not yet developed standardized procederus and pipelines within this domain and competence for these methods is still not widely available In this symposium we will assemble some of the leading proponents in big data neuroimaging approaches from Germany that will report on the state of the art, on their successes, but also failures, and which will present food for thought for the future prospects and challenges of big data approaches in and for psychiatry.