The transient uplift and subsidence of the Earth’s surface induced by mantle convection (dynamic topography) leaves an imprint on the stratigraphic record at inter-regional scales. Dynamically uplifted continental regions result in widespread erosional/non-depositional environments (sedimentary hiatus), while subsided regions result in continuous sedimentation. Thus, by mapping hiatus and no-hiatus signals on inter-continental scales, one gains a proxy for the long-wavelength uplift and subsidence associated with dynamic topography. In this contribution, we report on the use of hiatus maps as a constraint on mantle circulation models (MCMs), which make predictions of the history of dynamic topography. In order to make such a comparison, we filter the modelled dynamic topography through the available data points from the real maps to form hiatus/no hiatus signals. The resulting synthetic hiatus maps are then directly comparable to the true maps. By generating synthetic hiatus maps for a variety of high-resolution TERRA MCMs, we show that such maps allow for the falsification or verification of MCMs based on their prediction of dynamic uplift/subsidence events. We additionally find that eustatic sea-level variations are clearly highlighted by geological series in which the global ratio of hiatus/no-hiatus surfaces is significantly over-/under-predicted by the synthetic maps. We stress that, while plume histories in MCMs are constrained only by the surface tectonic history, this form of comparison paves the way for the validation of adjoint geodynamic models in which plume histories are constrained by seismic tomography.