The tectonically stable Bermuda Carbonate Platform has undergone repeated emersions and submersions in the course of the Pleistocene sea-level changes. Shallow seismic reflection profiling and high-energy vibration coring gave access to Pleistocene carbonates that are presently situated below sea level. A study of their diagenesis was undertaken in order to document the changes occuring in various diagenetic environments, to check a, possible correlation between the state of diagenetic preservation and stratigraphic age, and to estimate the odds for a complete diagenetic imaging of a sediment's pore-water history. For the purpose of comparison, several samples from the modern Bermuda Islands were analysed as well.As indicated by their biogenic constituents, the submarine Pleistocene carbonates were chiefly composed of aragonite and high-Mg-calcite at the onset of diagenesis, whereas they now commonly appear as mixtures of aragonite and low-Mg-calcite. Marine cements (five kinds of high-Mgcalcite cement, three kinds of aragonite cement) are, as a rule, restricted to the intraparticle pore space. Meteoric influence inflicted neomorphism, cementation with low-Mg-calcite (several kinds of cement) and solution on the marine carbonates. These are characterized by patchy lithification, abundant preserved pore space and incomplete mineralogical stabilization. Furthermore, especially in lagoonal sediments, burrow or root structures have been preferentially cemented. Indicators of karst and calichification are common. High-Mg-calcite is usually the first mineral to disappear under meteoric influence; aragonite is more stable. Intensely altered carbonates consist entirely of low-Mg-calcite, whereas dolomite does not occur at all.Diagenetic indications of meteoric-phreatic alteration are generally uncommon, but support the idea of fresh-water lenses that are largely restricted to the proximity of the present islands. Most of the carbonates, however, have received their decisive diagenetic imprints in the meteoricvadose environment.The diagenetic record within the carbonates is incomplete. Only one pore-water change is recorded, from a marine to a meteoric environment. Undoubted post-meteoric marine cements are absent from the studied Pleistocene material except for some intensely Iithified littoral deposits on the present islands. Obviously, the diagenetic potential of the marine environment was too small to compete with the meteoric environment. Under these circumstances, a reconstruction of pore-water history based on diagenetic sequences is deceptive.