The Bohemian Cretaceous Basin combines features of a shallow-water (mostly < 100 m) epicontinental seaway formed during a global transgression with those of a tectonically active, transtensional setting. The basin formed under a greenhouse climate and was affected by strong axial currents. Dense well-log coverage, combined with locally high-quality exposures and biostratigraphic control, make it possible to examine in three dimensions the geometries of genetic sequences and interpret their controlling variables. Sand-dominated deltas formed sequences at several spatial scales that reflect nested transgressive–regressive cycles with durations ranging from tens of thousands of years to millions of years. Progradation directions and distances, thicknesses and internal geometry of the individual sequences were controlled primarily by intrabasinal faulting, basin-scale changes in subsidence rate, eustatic fluctuations and localized bathymetric changes due to successive filling of the basin. Along-strike change in sediment input from different parts of the source area and a short-lived uplift of a secondary clastic source provided additional controls on the sequence geometry. Efficient hypopycnal transport combined with redeposition of fine clastics in shallow water promoted development of steep slopes of sand-dominated deltas while preventing downlap of muddy clinoforms; most of the suspended load became deposited downcurrent in subhorizontal or gently dipping bottomsets. Long-term accommodation rates were low during the Early to Middle Turonian, with minor intrabasinal faulting, but became accelerated in the Late Turonian and Early Coniacian. This acceleration was caused at least partly by increased subsidence rate accompanied by structural partitioning of the depocentre and partly compensated by increased sediment input indicating increased uplift rates in the Western Sudetic Island source area. This event probably reflected an increase in the regional strain rate in Central Europe. The succession of two major flooding events in the Early Turonian and late Early Coniacian, separated by a low-accommodation interval in the Middle Turonian, shows a close similarity to published estimates of long-term eustatic curves. However, the eustatic component of accommodation rate in the Bohemian Late Turonian and Coniacian is difficult to separate from accelerated subsidence. In several cases, evidence for short-term (100 kyr scale) forced regressions, independent of basinal structural activity, suggests small-scale eustatic falls at rates which, as presently understood, cannot be explained other than by a glacio-eustatic mechanism.