A series of real-data integrations of the National Center for Atmospheric Research Community Climate Model with tropical heat anomalies display regions of pronounced subsidence and drying surrounding the anomaly. The present emphasis is upon subsidence and drying centers located several thousand kilometers westward and poleward of the heating. These features are repeatedly found in several different series of medium to extended range forecast experiments, including cases of tropical Atlantic heating and tropical east Pacific heating. This highly predictable sinking response is established within the first five days of these integrations. The normal modes of a set of primitive equations linearized about a resting basic state are used to partition model response into gravity-inertia and Rossby modes. The results show that most of the vertical motion response can be explained by gravity-mode contributions. The sensitivity of the response is examined through a series of numerical experiments with a simple global forecast model. These integrations suggest that the subsiding response surrounding the heated region is somewhat sensitive to the ambient circulation. In particular, the extratropical response tends to be greatest in the winter hemisphere, and it is relatively less sensitive to the precise location of the tropical heating than to the nature of the zonally averaged background flow. Further experimentation suggests that the peak subsidence response is almost linear in the heating amplitude. These experiments also demonstrate that a significant portion of the early response occurs independently of any fluctuations of the vorticity field and therefore is not merely a secondary circulation associated with extratropical Rossby wave responses. The latter response is relatively more sensitive to the presence of longitudinal vorticity gradients, and the dynamical interpretation is then less clear.