Abstract: A new time-extended rotation model is developed used to construct data warehouse, with some related concepts elucidated covering abstract attribute, generalized abstract attribute, time-labeling, time domain and widened abstract attribute. The principle, classification and correlated factors of time-extended rotation model are described, with the subject-oriented and time-oriented methods also introduced as the construction strategies. With some application examples, the advantages and disadvantages of traditional DW model and time rotation model are discussed in view of the storage efficiency, data extraction efficiency, convenience in construction and practicability. The data warehouse constructed using the time rotation model will support the OLAP and DM well.