Phenology of trees of temperate regions has become an important research topic for global change issues over the last two decades because of its major role in regulating ecosystems processes such as carbon cycle and species interactions. While spring phenology has often been studied, autumn phenology is still poorly understood. Most of the studies focus on specific phenophases, such as budburst or flowering, and generally do not consider the carry over effect of a phenophase on following ones during the same annual cycle (e.g. budburst and bud set). Using beech saplings in a reciprocal transplanting experiment between two elevations in the Swiss Jura, we hypothesized that: (i) cooler or warmer winter conditions would affect the following spring and autumn phenophases, as well as the spring phenology of the following year. (ii) These changes could be related to non-structural carbohydrates (NSC) storage and/or to a shift in the dormancy release driven by chilling and forcing temperature requirements. Our results showed that the advance and the delay of budburst timing induced by warmer and cooler winter conditions, respectively, has impacted the following autumn phenology: an advanced spring budburst has led to an earlier autumn bud set. This suggests that the potential delay in senescence processes due to global warming might be smaller than expected and that the assumption of longer growing season does not necessarily apply. The dependency of these two phenophases might be explained by the building up of the NSC storage. Furthermore, no carry over effect was found between the advanced or the delayed budburst timing in 2014 and the spring phenology of 2015, indicating that the following year, the current climatic variables were the dominant drivers of spring phenology. Adapting phenological models to the whole annual phenological cycle would improve predictions of tree phenology under future climate warming conditions.