It is argued that, because of inherent weaknesses in design and procedure, the U.S. National Toxicology Program study of the carcinogenicity of inhaled vanadium pentoxide does not provide adequate evidence to support the classification by regulatory authorities of vanadium pentoxide as a Group 2B (possible) human carcinogen. The extension by one regulatory authority of the carcinogenicity classification for vanadium pentoxide to cover all vanadium compounds is also questioned. Such an extension implies that the toxic effect of some unknown vanadium species is more powerful than that of any oxygen species generated from the oxygen atoms in vanadium pentoxide, and that vanadium in any form can be converted in vivo to an undefined toxic species. There is no experimental or theoretical basis supporting this hypothesis. For oxygen-containing compounds like vanadium pentoxide, there is a need for some form of toxicity classification related to their oxygen content since it is likely to be the most reactive component of such compounds. For all particulates, such as those of crystalline vanadium pentoxide, the special toxicity of particulates, and especially nanoparticles, also needs separate consideration and classification.