In a recent paper, the author demonstrated that, in contrast with the prevailing view of eventual gradual regional differentiation from a hypothetical Cretaceous pantropical mangrove belt around the Tethys Sea, the Caribbean mangroves originated de novo in the Eocene after the evolutionary appearance of the first mangrove-forming tree species known for the region, the ancestor of the extant Pelliciera. This paper represents a second step in the analysis of the evolution of Caribbean mangroves dealing with the most important change experienced by these communities, occurring across the Eocene‒Oligocene transition (EOT), which is termed here the Caribbean mangrove revolution. This shift consisted of the disappearance of the primeval Pelliciera mangroves and their replacement by mangrove communities dominated by Rhizophora, a newly emerged mangrove tree that still dominates extant Caribbean mangroves. This paper first reviews the available literature on the EOT global disruption (tectonic and paleogeographic reorganizations, ocean circulation, cooling, Antarctic glaciation, sea-level fall) and its regional manifestations in the study area, along with the corresponding biotic responses. This provides the paleoenvironmental framework with which to analyze the EOT mangrove revolution using the nearly 80 pollen records available for the region. In the circum-Caribbean region, cooling of 3-6 °C and a sea-level fall of 67 m were recorded between 33.8 and 33.5 Ma, which led to significant shifts in dispersal pathways and barriers, as well as in marine paleocurrents. Late Eocene mangroves were dominated by the autochthonous Pelliciera (up to 60% of pollen assemblages), while Rhizophora, which likely arrived from the Indo-Pacific region by long-distance dispersal, was absent or very scarce. After the EOT, the situation was radically different, as the mangroves were widely dominated by Rhizophora, and Pelliciera, when present, was a subordinate mangrove element (<10%). At the same time, Pelliciera, which had been restricted to a small patch (Central America and NW South America or CA/NWSA) during the Eocene, expanded its range across the Caribbean and beyond, always as a minor component of Rhizophora mangroves. The dominance shift could have been due to the cooling, by favoring the expansion of the euryclimatic and vagile Rhizophora over the stenoclimatic Pelliciera, of limited dispersal ability. This is considered a case of competitor coexistence by niche segregation. In addition, Rhizophora could have facilitated the expansion of Pelliciera by providing refuge against environmental and biotic stressors, notably light intensity and salinity. The Eocene Pelliciera mangroves never returned, but this species survived to the present as a minor element and experienced significant range shifts along three main phases, namely, EOT–Miocene expansion to the whole neotropics, Mio-Pliocene contraction to the southern Caribbean margin and Pliocene to recent reorganization to the original Eocene CA/NWSA location. The potential role of Neogene and Pleistocene climatic shifts and human activities in these biogeographical loops (taxon cycles) is discussed, with an emphasis on precipitation. The paper ends by suggesting some prospects for future research.