Remote sensing is a modern geo-informatics technology with a high priority for practical applications. However, theoretical studies in this field are relatively not updated compared with its fast-expanding technical applications. Thus, establishing new theories to support further development and deep applications of remote sensing is important. This paper focuses on new theoretical exploration on the inherited cross time and space features in general remote sensing applications to guide future remote sensing applications.The historic development of remote sensing technology is reviewed, and existing composition and classification of remote sensing are analyzed. Then, the demand for a new classification method of remote sensing is examined.Cross time and space features commonly found in remote sensing applications are evaluated, and a new vision toward remote sensing applications is discussed. A new theory isproposed.A new method ology-based classification approach is proposed for remote sensing. This technique can provide the foundation for some new research directions in remote sensing.Cross time and space Remote Sensing(CRS) theory is proposed,and its basic framework is described to provide a new vision toward future remote sensing applications. Three typical examples are used to illustrate the actual application of CRS theory in our previous study.CRS theory and its application are developed because of the strong demand for further development and applications of remote sensing. This study will serve as apractical guide for various future applications of this technology.