This paper introduces the concept of the two-phase seamless integration model (2P-SIM), which is a general model to describe and solve the seamless integration problem of spatial data in the distributed heterogeneous environment. The first phase of this model deal with the classify progress; after this phase spatial data participate in seamless integration is split into respective groups, each group corresponding a computation unit in the second phase. The second phase of the model can be split into three correlative parts, to deal with the spatial geometry seamless integration, the spatial object attribute seamless integration, and the accuracy fusion principle respectively. In each part, there are some elements or factors to describe and define various aspects of the seamless integration requirements and process manner; each element can select one of some predefined algorithms from the model, these algorithms are also illustrated and discussed.