Tree-related microhabitats (TreMs) provide a quantitative indicator of habitat heterogeneity in forests, including beech (Fagus) forests. However, systematic analyses of the factors driving TreM diversity and composition in Fagus sylvatica and F. orientalis forests are lacking. In this study, the TreMs of beech forests on 203 plots of 22 forest sites (production and old-growth forest) across the full longitudinal range of both species were assessed following a standardized TreM protocol. A unified diversity and ordination framework based on Hill numbers was applied to account for unobserved TreM types and to extend the sensitivity of our findings focusing from rare to dominant TreMs. The composition of TreM assemblages was mostly determined by Fagus species and elevation, a surrogate for climate, and with focus on dominant TreMs by DBH, whereas old-growth versus production forest had no effect. The coverage of detected TreMs per plot increased with the number of trees assessed and DBH, but was lower in old-growth than in production forests. When standardized for sampling coverage, the diversity of rare and dominant TreM types was higher in old-growth than in production forests, but increased with elevation only focusing on dominant TreMs. These findings corroborate regional studies showing a higher TreM diversity in old-growth forests. Moreover, they demonstrate the importance of focusing conservation efforts on forests of both Fagus species and at different elevations, covering the full range of TreM assemblages. Future studies comparing TreM diversity in different forests should standardize diversity by sample coverage, as currently done in many biodiversity studies.