We evaluated ages of larvae in southern Appalachian populations of the salamanders Gyrinophilus porphyriticus and Pseudotriton ruber by the method of skeletochronology applied to femurs. In P. ruber, the method yielded a strong correlation between snout–vent length and age and provided estimates in agreement with earlier studies of larval growth and age that were based on size-frequency data. However, in G. porphyriticus, the correlation between snout–vent length and age was weak and nonsignificant. This result was in accordance with previous studies that showed that larvae of this species do not sort by size into discrete age classes. The failure of skeletochronology in G. porphyriticus was attributed to either extreme variation in growth rate or to some deficiency of the methodology itself when applied to this species. We suggest that larvae of G. porphyriticus may not develop well-defined annual bone layers because of their adaptation to relatively aseasonal, subterranean microhabitats in headwater springs. In contrast, larvae of P. ruber, although overlapping in macrohabitat with G. porphyriticus, are essentially surface-dwellers that are exposed to the seasonal climatic fluctuations of the region.