Nowadays, biologists use a number of large biological databanks to find relevant information for their research. Users of these databanks face a number of problems. One problem is that users are required to have good knowledge about the contents, implementations and conceptual models of many databanks to be able to ask precise and relevant questions. Further, the terminology that is used in the different databanks may be different. Also, when asking complex queries to multiple databanks, users need to construct a query plan on their own possibly leading to poor performance or not even obtaining results. To alleviate these problems we define an architecture for systems that deal with these problems by allowing for a transparent and integrated way to query the multiple sources. The contribution of this paper is threefold. First, we describe a study of current biological databanks. Then, we propose a base query language that contains operators that should be present in any query language for biological databanks. Further, we present an architecture for a system supporting such a language and providing integrated access to the highly distributed and heterogeneous environment of biological databanks.