The paleobotanical database for the Cenozoic terrestrial vegetation of the Greater Antilles consists of the Saramaguacan (Cuba) and Chapelton (Jamaica) floras of middle Eocene age; the San Sebastian (Puerto Rico) flora of middle Oligocene age; the Yumari (Cuba) and Artibonite (Haiti) floras of Miocene to Mio-Pliocene age; two small floras of unspecified Tertiary age (Dominican Republic, Haiti); and three Quaternary floras from Cuba, Puerto Rico, and Haiti. This meager basis is not adequate to trace the environmental and vegetational history of the region but supplemental information from calcium-selenium geochemistry and oxygen isotopes document tropical conditions and vegetation through the middle Eocene with geographic affinities primarily with North America, and increasing representation of South American elements and temperate species after the global temperature declines of the middle Eocene and the early Miocene. The geohistory of the Caribbean islands is still unsettled and this history is directly relevant to assessing the relative importance of vicariance and dispersal in populating the islands. The trend toward models depicting less extensive movement of the land fragments, less direct association with continental land masses, fragmented continental connections, inundations, and the original appearance of many as islands supports dispersal as an important means of distribution. Collisions, extinctions, an unsettled geohistory, and the few area cladograms available for plant taxa complicate assessing the relative role of vicariance, but zoogeographic studies indicate it was also an important means of distribution and speciation. Achieving an accurate understanding of the immensely complex history of the Greater Antillean biota is deferred by efforts to promote one mechanism at the expense of the other.