Abstract Earlier spring growth onset in temperate forests is a visible effect of global warming that alters global water and carbon cycling. Consequently, it becomes crucial to accurately predict the future spring phenological shifts in vegetation under different climate warming scenarios. However, current phenological models suffer from a lack of physiological insights of tree dormancy and are rarely experimentally validated. Here, we sampled twig cuttings of five deciduous tree species at two climatically different locations (270 and 750 m a.s.l., ~ 2.3 °C difference) throughout the winter of 2019–20. Twig budburst success, thermal time to budburst, bud water content and short-term 2H-labelled water uptake into buds were quantified to link bud dormancy status with vascular water transport efficacy, with the objective of establishing connections between the dormancy status of buds and their effectiveness in vascular water transport. We found large differences in the dormancy status between species throughout the entire investigation period, likely reflecting species-specific environmental requirements to initiate and release dormancy, whereas only small differences in the dormancy status were found between the two studied sites. We found strong 2H-labelled water uptake into buds during leaf senescence, followed by a sharp decrease, which we ascribed to the initiation of endodormancy. However, surprisingly, we did not find a progressive increase in 2H-labelled water uptake into buds as winter advanced. Nonetheless, all examined tree species exhibited a consistent relationship between bud water content and dormancy status. Our results suggest that short-term 2H-labelled water uptake may not be a robust indicator of dormancy release, yet it holds promise as a method for tracking the induction of dormancy in deciduous trees. By contrast, bud water content emerges as a cost-effective and more reliable indicator of dormancy release.< Réduire