An improved end-use water demand model was developed. The model intends to predict at multiple spatial and temporal scales. The model considers known water demand determinants and unexplained behavior. The model can be used to predict peak water demand for residential water end-uses. The paper also assesses the impacts of structural and demographic factors.