A simple model is used to study the mechanisms which control the local and remote (teleconnection) response of the atmosphere to the thermal forcing resulting from sea surface temperature (SST) anomalies located at various latitudes. The model chosen is a linear baroclinic spherical primitive equation model containing a zonally symmetric basic state with horizontal and vertical shear. An iterative procedure is developed in which the total diabatic heating resulting from the initial heating by the SST anomaly is calculated via feedbacks between the heating and the dynamic response of the system. Depending on the latitudinal location of the SST anomaly, two major limits of atmospheric response may be identified. The first, the “diabatic limit”, occurs with the SST anomaly embedded in weak low-latitude basic flow and results in a strong enhancement of the initial anomaly response through a vigorous positive dynamics-diabatic beating feedback. Strong teleconnections are evident between low and high latitudes. The second domain, the “advective limit”, occurs when the SST anomaly is placed at higher latitudes in the vicinity of the westerly maximum. The local response is extremely small due to the creation of an indirect zonal circulation in the vicinity of the anomaly which is related to the strength of the local basic flow and the latitude of the forcing due to rotational limitations on the relative scale of the vertical velocity. In contrast to the diabatic limit, the form of the principal forced mode appears unimportant in determining the final response. That is, the dynamics-diabatic heating feedback is weak and only marginally positive. The form of the remote high-latitude response in all cases is scale selective and only the largest scale transmitted modes are excited. It is argued that these are closest to resonance in latitudes of strong basic zonal flow. The remote response shows a distinct structural difference on either side of the westerly maximum, being highly baroclinic on the equatorial side but barotropic on the polar side. Limited cross-equatorial propagation occurs due to the existence of critical latitudes on the equatorial side of the forcing. The model results are used to interpret the experimental results obtained from general circulation models (GCM) and provide a rationale for the existence of teleconnections found between low and high latitudes when SST anomalies were imposed in the equatorial oceans. Furthermore, the results suggest why “super-anomalies” were required in midlatitudes in some GCM experiments in order to produce a response which was measurable above the noise level of the model. It is shown that it is possible to resolve the apparent paradox between the minimal response of GCMs to the imposition of middle latitude SST anomalies and the observations of Namias (1976a) and Davis (1978) who related atmospheric anomalies at least relative to summer SST anomalies. It is argued that only at times of small basic flow (i.e., summer) will a significant response arise in midlatitudes. Finally, the relevancy of the model results to such features as the South Pacific cloud band and the Southern Oscillation is discussed.