The poleward transport of heat in the mid-Cretaceous (100 Ma) is examined using an idealized coupled ocean-atmosphere model. The oceanic component consists of two zonally averaged basins representing the proto-Pacific and proto-Indian oceans and models the dynamics of the meridional thermohaline circulation. The atmospheric component is a simple energy and moisture balance model which includes the diffusive meridional transport of sensible heat and moisture. The ocean model is spun up with a variety of plausible Cretaceous surface temperature and salinity profiles, and a consistent atmosphere is objectively derived based on the resultant sea surface temperature and the surface heat and freshwater fluxes. The coupled model does not exhibit climate drift. Multiple equilibria of the coupled model are found that break the initial symmetry of the ocean circulation; several of these equilibria have one-cell (northern or southern sinking) thermohaline circulation patterns. Two main classes of circulation are found: circulations where the densest water is relatively cool and is formed at the polar latitudes and circulations where the densest water is warm, but quite saline, and the strongest sinking occurs at the tropics. In all cases, significant amounts of warm, saline bottom water are formed in the proto-Indian basin which modify the deepwater characteristics in the larger (proto-Pacific) basin. Temperatures in the deep ocean are warm, 10¿--17 ¿C, in agreement with benthic foraminiferal oxygen isotope data. The poleward transport of heat in the modeled Cretaceous oceans is larger than in some comparable models of the present day thermohaline circulation and significantly larger than estimates of similar processes in the present-day ocean. It is consistently larger in the polar sinking cases when compared with that seen in the tropical sinking cases, but this represents an increase of only 10%. The largest increase over present-day model transports is in the atmospheric latent heat transport, where an increased hydrological cycle (especially in the tropical sinking cases) contributes up to an extra 1 PW of poleward heat transport. Better constraints on the oceanic deepwater circulation during this period are necessary before the meridional circulation can be unambiguously described. ¿ American Geophysical Union 1996