The lack of annual growth rings in the majority of tropical tree species greatly limits our understanding of the long-term dynamics of tropical forests. To address this problem, several methods have been developed to estimate the age of tropical trees from diameter growth data. These past approaches, however, suffer from two major flaws: (1) they assume a deterministic age–size relationship for a tree species, and (2) they have not been verified with independently derived age data. In this paper, I present a new approach that uses diameter growth rates, independent of tree size, that are stratified by crown class to estimate the age of individual trees. Past approaches have assumed that when present-day canopy trees were juveniles they grew at rates similar to conspecifics currently in the understory. In contrast, the crown class model assumes that present-day canopy trees have grown at rates similar to conspecifics in the same crown class, irrespective of size, through-out ontogeny. The crown class model was compared to a periodic annual increment (PAI) model typical of past approaches in mixed oak-hardwood forest in the southern Appalachians, USA. Tree ages were obtained independently from tree cores from three stands of differing age structure. Comparisons between the two models were made for species' population age structures (independent of stands) and stand age structures (independent of species). Age estimation errors for the crown class model were lowest for relatively shade-intolerant species such as yellow poplar (Liriodendron tulipifera) and chestnut oak (Quercus prinus), but increased with the increasing shade tolerance of the species. The PAI model followed the opposite pattern, providing the most accurate age estimates for shade-tolerant species. These results were consistent with the underlying ecological assumptions of each model. When predicted age distributions for individual stands were compared to the true age distributions, the PAI approach had higher estimation errors than the crown class model in almost every case. In addition, the PAI model had a strong tendency to overestimate tree ages. A mixed model that used crown class age estimates for shade-intolerant species and PAI model age estimates for shade-tolerant species generated the most accurate esti-mates of stand age distribution. The results of this study suggest new opportunities for the study of long-term dynamics in tropical forests and underscore the importance and utility of validating models with independent data.