Climate affects tree growth but the effect size can be modulated by other variables, including tree’s age and size. To assess how climate sensitivity changes over the life of a tree, previous studies mostly stratified trees into age classes, while cambial ring-age stratification (age-band decomposition) was less frequently used. However, trees do not age as other organisms and arguably age is mainly a proxy for size, which in contrast to age has been shown to affect wood anatomy and physiology. Stem diameter classes, calculated from cumulative ring width, could thus facilitate a more direct assessment of size effects. Here we compare these three methods, which differ regarding how they stratify data into age/size classes. We found that using age-band decomposition and cumulative ring-width classes had major advantages over the tree-age method: (a) age and size are decoupled from other temporal changes, like atmospheric CO2 concentration or nitrogen deposition, which excludes potential biases. (b) Shifts in climate sensitivity occur earlier than estimated by the tree-age method. (c) Younger/smaller classes can be assessed. Furthermore, direct comparison supports that size, rather than age, alters climate sensitivity. Therefore, the cumulative ring-width method appears to be the best approach to assess the effect of ontogenetic changes on a tree’s climate sensitivity. Understanding how climate sensitivity changes when trees get older and larger is important for forest ecology and management, climate reconstructions, global carbon models and can help to study age and height limitations of trees.