For at least five million years, the motion of the subducting Pacific plate has been steady in a direction approximately perpendicular to the strike of the island arc in northeast Japan. We have shown that the present-day topography, gravity and seismicity patterns can be modeled successfully by assuming an approximately steady tectonic compressive driving force and a plausible crustal rheology and by using a finite element modeling code which allows stick-slip faulting, erosion/deposition and time-dependent loading.In this paper we show how the strength of interplate coupling with depth and along strike is constrained by the observed topography and gravity, and we examine the correlation of coupling with observed interplate and intraplate seismicity. Topography and the gravity residual anomaly in the eastern coastal area are significantly higher in region 1 (the Kitakami range and off Sanriku) than in region 2 (off Miyagi). In order to match the observed topography and gravity, the strength of interplate coupling in region 1 is about two to four times greater than in region 2, and there is about twice as strong a coupling at greater depths. Coupling in region 1 and region 2 are considered as two extreme cases. In region 3 (the Abukuma range and off Fukushima), the coupling strength is between that of region 1 and region 2.The model predicts a higher potential for earthquakes in the crust and mantle wedge apex for the region with stronger coupling, region 1, which is consistent with the observed intraplate seismicity. Observed patterns of interplate seismicity are also consistent with differences in interplate coupling; the weaker coupling in region 2 indicates a lack of seismogenic potential which leads to a low probability for large earthquakes in that region.