Accurate plant phenology (seasonal plant activity driven by environmental factors) models are vital tools for ecosystem simulation models and for predicting the response of ecosystems to climate change. Since the early 1970s, e!orts have concentrated on predicting phenology of the temperate and boreal forests because they represent one-third of the carbon captured in plant ecosystems and they are the principal ecosystems with seasonal patterns of growth on Earth (one-"fth of the plant ecosystems area). Numerous phenological models have been developed to predict the growth timing of temperate or boreal trees. They are in general empirical, nonlinear and non-nested. For these reasons they are particularly di$cult to "t, to test and to compare with each other. The methodological di$culties as well as the diversity of models used have greatly slowed down their improvement. The aim of this study was to show that the most widely used models simulating vegetative or reproductive phenology of trees are particular cases of a more general model. This uni"ed model has three main advantages. First, it allows for a direct estimation of (i) the response of bud growth to either chilling or forcing temperatures and (ii) the periods when these temperatures a!ect the bud growth. Second, it can be simpli"ed according to standard statistical tests for any particular species. Third, it provides a standardized framework for phenological models, which is essential for comparative studies as well as for robust model identi"cation.