It has been revealed in previous studies that one possible mechanism responsible for the interdecadal change of the Northern-Hemisphere (NH) circulation is the interdecadal sea-surface temperature (SST) change. It is thus hypothesized that the interdecadal SST change is one of the possible mechanisms to induce the interdecadal change of the NH winter circulation. To test this hypothesis, two multi-decade (1946-92) climate simulations were performed with the R15 resolution of the NCAR CCM1: one incorporating the 12 calendar month climatological SSTs, the other using the real-time SSTs. By contrasting the results of these two climate simulations, the effect of SST anomalies on the interdecadal change of the NH winter circulation was found to induce (1) the equatorward expansion of the circumpolar vortex, (2) the deepening of three climatological troughs, and (3) three interdecadal variation modes of stationary eddies: Pacific/North America (PNA), PNA west (PNAW), and North Atlantic (NA) modes. These three interdecadal modes were found to be equivalent barotropic in their vertical structure. Horizontally, both the PNA and PNAW modes exhibited a teleconnection pattern over the Pacific/North American region, while the NA mode possessed a north-south three-cell structure over the Greenland/North Atlantic region. The temporal variations of these three modes consisted of a decadal trend and 15-20 year low-frequency oscillations;The interdecadal variation of general circulation statistics is regulated by that of the atmospheric circulation. In the thermal field, the transient heat flux diverges out of the warm anomalies and converges toward the cold anomalies. In the dynamic field, the cyclone activity over the midlatitude storm track region is affected by the variability of the north-south wind shear which is induced by the equatorward expansion of the circumpolar vortex;It is further demonstrated by linearized model simulations that (1) the tropical heating anomalies dominate over the extratropical heating anomalies as the primary mechanism to maintain the interdecadal change of the NH winter circulation, (2) the transient forcing anomalies are not important to the interdecadal circulation variability, and (3) the effect of the polar cooling is to amplify the stationary eddies.