Poleward transport of heat by the atmosphere as simulated by a recently developed general circulation model (GCM) is consistent with earlier GCM studies in being significantly greater than indicated from radiosonde (weather balloon) observations. However, total heat transport by the combined ocean-atmosphere system appears to be approximately the same in the models and in satellite observations of irradiances at the top of the atmosphere: in the models most of this transport takes place in the atmosphere whereas the combined satellite and radiosonde observations indicate that half or more of the transport takes place in the oceans. It is argued here that the atmospheric model results are generally closer to the truth than the radiosonde observations. If this is true, then the oceans transport less heat than often supposed, and conventional ideas about the genesis of climatic change through alterations in oceanic heat transport must be reexamined.