The results of genome-based research are increasingly becoming part of planning for care. The question remains whether the results of this research are representative, since it often excludes people with psychiatric conditions. Despite best intentions, in our desire to protect the vulnerable, we risk creating a situation where they are discriminated against in a double sense: non-inclusion to protect them from the risks of research leads to biased results, that in turn could lead to a lower standard of care in the future.Large-scale population biobanking projects link genetic data with the information participants provide regarding their health status, lifestyle (including alcohol consumption, smoking, etc.) and known environmental factors. The collection of samples and data constitutes the basis for future research projects on two levels. First, some diseases or phenotypes are collected with great precision during the recruitment for existing research protocols. Second, follow up projects are planned based on the analysis of data collected from questionnaires and analysis of biological specimens on the basis of identified phenotypes. Often, these data make it possible to assess and identify diseases ranging from diabetes to depression. Once a condition is identified, researchers need to collect further phenotypes or dig deeper into a recognized pattern.Commonly, participants in population cohorts are recruited for studies of common diseases. For example, there is a large number of cohorts focused on cardiovascular and metabolic conditions. Recruitment of population cohorts often excludes incapacitated individuals and people who, for several reasons, cannot provide a full informed consent. This means minors are often excluded from population cohorts and (sometimes) instead collected separately in biobanks for children. Against this background, it becomes obvious that there is a problem of representativeness for these categories of people that constitute an already known first bias in the collections. The reason for this is often administrative: handling issues of legal representatives may in fact prove to be too complicated on a population level.There is, however, a second layer of possible bias that arises in the planning of follow up research projects. Planning to re-call people with severe depression, addiction or initial stages of degenerative conditions (dementia, etc.), often entails several layers of ethical and logistical issues that do not arise in re-contact of participants with suspected diabetes or leukemia.If we use depression as an example, this problem becomes apparent. When people suspected to have depression are invited to participate in follow up research with a physician, they are asked to fill out questionnaires used to diagnose depression and identify suicide risks. Costly follow up visits with specialists are needed if a research participant discloses that he or she is contemplating suicide. For people with early stage dementia, the problem is more related to information and ability to pay attention, and the difficulties associated with explaining and understanding large amounts of complex information.This presentation will provide an overview of possible discriminatory patterns that jeopardize the possibility for psychiatric patients (in a very broad sense) to benefit from the potential of biobank-based research, with some case examples.