Two studies presented in the literature (Murray, Canned & Smith 1989; Hanninen 1991) evaluate the effect of increasing winter temperature on the probability of spring frost damage to trees, but yield contradictory results. It is unclear whether the disparity can be ascribed to the fact that different models were used, or is the result of different climatic warming scenarios being used, or is because the tree species at the different locations do indeed respond differently to warmer winters. To evaluate the effects of climatic warming to tree species in The Netherlands and in Germany, both models were fitted to long series of observations on the date of leaf unfolding of eleven tree species. The impact of the two scenarios (uniformly and non-uniformly changing winter temperature) on the date of leaf unfolding and on the probability of freezing temperature around that date was evaluated. To test the importance of adaptation to local climate, hypothetical provenance transfers were analysed. It was concluded that, for tree species in The Netherlands and Germany, the probability of spring frost damage will decrease. The contradictory results found in the literature could be ascribed to differences between provenances adapted to their local climate, and is not because different models and different climatic warming scenarios were used in these studies.