The discovery of new fossil bryophytes allows refined estimates of divergence times when the fossils have unambiguous synapomorphies enabling their confident assignment to nodes within molecular phylogenies. We use two fossil Radula species from Cretaceous age Burmese amber to estimate divergence times for Radula. One of these fossils, R. cretacea which has synapomorphies of subg. Odontoradula, has been used previously; the other, with relative synapomorphies of subg. Amentuloradula, has not. In combination these two fossils, when used to constrain the crown node age of their respective subgenera, result in median age estimates for the Radula crown node of 263 million years, under our preferred time-calibration scenario where subgeneric crown-node fossil assignments are coupled with a secondary maximum age constraint on the Porellales crown node. We explore other time-calibration scenarios, including deeper fossil assignments, and conclude that, under all, Radula as currently circumscribed is an outlier among land plant genera on the basis of its age. While this violates no established norms regarding how old genera should be, the absolute age, the relative ages of other families within Porellales, and the morphological distinctiveness of the two serial sister lineages, subg. Cladoradula, with seven species, and subg. Dactyloradula with one species, motivates our proposal to elevate these two subgenera to generic rank, resulting in three genera within the family Radulaceae. We provide diagnoses for genera, new combinations for species, and an emended circumscription for Radula, that reflect this change.