We evaluated the applicability of Rapoport’s Rule (RR) to the marine bony fishes of the world. The biogeographical pattern predicted by RR has been the subject of a large number of studies, some supporting it and some not. In this exercise, we attempted to generate results free of biases from taxonomy or geographic scale. The study area encompassed all world oceans. Our analysis was based on secondary data. We tested the relationship of the geographic range and the bathymetric range to the latitude gradient, using the method of Stevens (American Naturalist 1989, 133, 240–256). We compared all known species of marine bony fishes together, and performed a second analysis limited to the non-pelagic species. Our results were generated from a databank including 13,957 species. The results indicated that RR (latitudinal range) is valid for 11 regions. However, there were exceptions for the northern part of the Eastern Atlantic and Pacific, the northern part of the Western Indian, and the Arctic Oceans. The analyses of RR (bathymetric range) were supported for 13 regions. We found only one exception, for the northern part of the Western Indian Ocean. The applicability of RR and its relationship to energy supply, represented here by the temperature variability of the marine water masses, was clear. The exceptions appeared when some feature associated with the distribution of water masses was superposed on the latitudinal gradient, or when some life trait characteristic of a taxon, for instance body size, was confounded with the effect of climate variability.