Models used to predict the onset of fruit tree blossom under changed climate conditions should be physiologically based as much as possible. Pure optimized phenology models carry the risk of unrealistic predictions due to a misinterpretation of metabolic processes. This was the motivation determining the relevant phases for chill and heat accumulation, which induces cherry blossom (cv. Summit). Investigations are based on 8 years of observational and analytical data, as well as on controlled experiments. For ‘Summit’ buds, to be released from endodormancy, 43 chill portions from 1 September are necessary. After endodormancy release (t1), on average on 30 November, no further chilling is required, because no correlation between chill accumulation during ecodormancy and the subsequent heat accumulation until ‘Summit’ blossom exist. The declining amount of heat, which induces cherry blossom after t1—shown in several forcing experiments—seems to be the result of the declining bud’s abscisic acid (ABA) content, up to ~50% until the beginning of ontogenetic development. Shortly after t1, when the bud’s ABA content is high, a huge amount of heat is necessary to induce cherry blossom under controlled conditions. Heat requirement reduces during ecodormancy along with the reduction in the ABA content. According to these findings, plant development during ecodormancy is suppressed by low temperatures in the orchard and a slowly declining bud’s ABA content. These results should lead to a better consideration of the ecodormancy phase in phenology models.