The hypothesis that caves in the Florida Peninsula are tied to Quaternary sea levels was proposed by hydrogeologists, without data, some 40 years ago. The hypothesis is a version of glacial control of cave levels, which is the logical combination of the water-table theory of speleogenesis and the concept that base level positions the water table. At the USA type example of glacial control of cave levels—Mammoth Cave in the Paleozoic rocks of Kentucky—the intermediary is base level determined by rivers. By hypothesis, the intermediary for Florida is glacioeustatic sea level. This paper presents elevation data that supports this hypothesis. Recent cave surveys in the air-filled caves and spot elevations from archived maps reveal prominent levels of passages centered at 5, 12, 21, and 30 m above sea level over broad areas. They do not follow the large-scale structure of the Floridan aquifer. Instead, they align with nearby, coastal marine terraces identified as modal peaks on frequency plots from various topographic data bases. Levels matching with the three highest terraces—Wicomico, Penholoway, and Talbott—are particularly clear. Lower levels, if they accord with sea-level stands, are likely composites. Data from cavities encountered in drilled wells (e.g., bit drops) and spot elevations from archived underwater cave maps demonstrate passage levels at depths of 15, 30, 70, and 90-120 m below the modern water table. The depths below water table are similar to the depths below sea level of distant submerged terraces and paleoshoreline features identified using multibeam bathymetric data in the Gulf of Mexico. The cave, bit-drop, and terrace data are all consistent with the concept that Quaternary sea level is the fundamental control on the cave-scale porosity within the Floridan aquifer. This conclusion does not rule out the possibility that lithologically favored positions, paleokarst features and confining units, and mixing zones are also involved in the location of caves levels in this near-coastal environment.