The building stock of Dresden consists of a large number of brick buildings constructed during the 20th century. Different type of bricks were sampled and their moisture storage function was measured using pressure plate apparatus and desiccators with different saturated salt solutions. A multivariate cluster analysis procedure was carried out to detect natural groupings within the data. Multi-modal water retention functions were applied to smooth and interpolate between the scattered data of the discovered clusters to provide a complete, unbroken, and consistent functional description. In total, four representative retention transfer functions could be found for the collection of the bricks of the Dresden building stock.