As the original molluscan radula is not known from direct observation, we consider what the form of the original radula may have been from evidence provided by neomenioid Aplacophora (Solenogastres), Gastropoda, Polyplacophora, and the Cambrian fossil Wiwaxia corrugata (Matthews). Conclusions are based on direct observation of radula morphology and its accessory structures (salivary gland ducts, radular sac, anteroventral radular pocket) in 25 species and 16 genera of Aplacophora; radula morphogenesis in Aplacophora; earliest tooth formation in Gastropoda (14 species among Prosobranchia, Opisthobranchia, and Pulmonata); earliest tooth formation in four species of Polyplacophora; and the morphology of the feeding apparatus in W. corrugata. The existence of a true radula membrane and of membranoblasts and odontoblasts in neomenioids indicates that morphogenesis of the aplacophoran radula is homologous to that in other radulate Mollusca. We conclude from p redness of salivary gland ducts, a divided radular sac, and a pair of anteroventral pockets that the plesiomorphic state in neomenioids is bipartite, formed of denticulate bars that are distichous (two teeth per row) on a partially divided or fused radula membrane with the largest denticles lateral, as occurs in the genus Helicoradomenia. The tooth morphology in Helicoradomenia is similar to the feeding apparatus in W. corrugata. We show that distichy also occurs during early development in several species of gastropods and polyplacophorans. Through the rejection of the null hypothesis that the earliest radula was unipartite and had no radula membrane, we conclude that the original molluscan radula was similar to the radula found in Helicoradomena species. J. Morphol. 257:219–245, 2003. © 2003 Wiley-Liss, Inc.