A Tree related Microhabitat (TreM) is a distinct, well-delineated morphological singularity occurring on living or standing dead trees, that constitutes a particular and crucial substrate or life site for species. TreMs are essential to support forest biodiversity. The retention of some trees bearing TreMs is increasingly prescribed in sustainable forest management guidelines. However, this may not be sufficient to ensure an appropriate density and diversity of TreMs to support TreM-dwelling taxa over time. Indeed, trees that are harvested often either display TreMs or have the potential to develop them over time (future habitat trees). Moreover, little is known about the dynamic process of TreM’s formation at the tree level. Most of the TreM data currently available are cross-sectional data (trees are observed only once) making it more difficult to estimate TreM formation rates. In this presentation we show how to estimate the probability of a new TreM being formed during tree growth by adapting statistical survival analysis approaches. Using a harmonized European TreM database, we analyze variations in the probability of formation of a new TreM on a tree of a given DBH for 11 different groups of TreMs and different groups of tree species. Then, using Samsara, a spatially-explicit individual-based forest stand model, we show how thinning regimes influence the TreM profile of uneven-aged mixed stands of spruce, fir and beech over time.