Coseismic strain is fundamentally important for understanding crustal response to changes of stress after earthquakes. The elastic dislocation model has been widely applied to interpreting observed shear deformation caused by earthquakes. The application of the same theory to interpreting volumetric strain, however, has met with difficulty, especially in the far field of earthquakes. Predicted volumetric strain with dislocation model often differs substantially, and sometimes of opposite signs, from observed coseismic volumetric strains. The disagreement suggests that some processes unaccounted for by the dislocation model may occur during earthquakes. Several hypotheses have been suggested, but none have been tested quantitatively. In this paper we first examine published data to highlight the difference between the measured and calculated static coseismic volumetric strains; we then use these data to provide quantitative test of the model that the disagreement may be explained by the change of pore pressure in the shallow crust. The test allows us to conclude that coseismic change of pore pressure may be an important mechanism for coseismic crustal strain and, in the far field, may even be the dominant mechanism. Thus in the interpretation of observed coseismic crustal strain, one needs to account not only for the elastic strain due to fault rupture but also for the strain due to coseismic change of pore pressure.