Assessing tree ages is important for the understanding of forest recruitment patterns and tree growth. However, little experience exists in the subtropics and accurate aging usually involves potentially damaging techniques such as tree coring or even the use of entire cross sections, which is not recommendable in endangered species or ecosystems. We provide an example of how age may be predicted on the basis of tree size and site conditions, using Polylepis australis of Central Argentina. Our study was conducted in two entire river basins where 96 independent trees were successfully cored for age analysis. Site and tree characteristics were registered and tree age determined through standard dendrochronological techniques. The multiple regression procedure selected tree circumference and proportion of rock under the tree canopy as signiWcant variables explaining 48% of the variation in number of growth rings. Trees surrounded by rock clearly grew slower than those surrounded by a higher proportion of soil and vegetation. A comparison with a one site study shows that the ability to predict P. australis age is clearly reduced when geographical range is extended but the explanatory power of our model is still high enough for certain applications and within the range of other one site studies. We suggest that tree size and site characteristics may be used to predict age in other subtropical mountain forests with a well marked growth season.