There is a long-standing debate on whether the occurrence of the iconic high-Andes Polylepis woodlands as small and isolated fragments is of natural or anthropogenic origin. We make inferences regarding the fragmentation history based on both a new population genetic study on P. besseri and a synthesis of available studies on the population genetics of Polylepis woodlands. We infer the timing of the main woodland fragmentation event by analysing: (1) the remaining levels of population genetic diversity and the relation to population size; (2) among-population genetic differentiation; and (3) the difference in genetic diversity between the offspring and adult generation. We retrieved seven publications on the population genetics of five Polylepis spp. We did not find a relationship between population size and genetic diversity, and genetic differentiation was low compared with that reported for similar plant species. These findings do not support a history of long-term fragmentation. The offspring showed a loss of genetic diversity and increasing differentiation compared with adults, suggesting that the main habitat fragmentation event is of relatively recent origin. For P. besseri, no significant differences were found between the adult and offspring genetic variation. We discuss the conservation and restoration consequences for this important high-Andean genus. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 544–554.