Under southern Uruguayan conditions (34°37'SL;56°11'O, 40 m) winter chilling requirements are not always satisfied for the main commercial apple (Malus domestica Borkh) cultivars. The use of chemical rest breaking agents (mineral oils and hydrogen Cyanamid) to compensate these requirements is widely used. Therefore winter chill quantification should be a relevant issue to adjust this technique. To do so, adequate models to represent the advances in the endodormancy process as a function of climatic conditions are necessary. The UTAH and CH models are usually applied, regardless of which species or cultivar is considered. Although these models have the advantage to be widely spread and incorporated to commercial production, there are evidences of the lack of adjustment to the phenologyc behaviour of the main cultivars. This work's goal was to quantify the occurrence of winter chill during the endodormancy period with six models, in four apple cultivars ('Red Chief', 'Granny Smith', 'Royal Gala' and 'Fuji') during 2005 and 2006. The models used were CH, UTAH, UTAH+, Dynamic Model, North Carolina Model and Model for Subtropical Conditions. The beginning of endodormancy release was determined when 50% of leaf fall was reached and the process finished when 50% of one-year-old shoot buds sprouted in a forcing chamber. All model quantifications were based on hourly temperature data, obtained from sensors located in the tree and at the nearest weather station. The greatest temperature amplitude registered at field conditions, compared to weather station conditions, modified up to 2.5 times, the quantifications performed by the same model, being the UTAH model, the most affected by this phenomenon. These results show differences among models and years, consequently, in order to determine the viability of them in southern Uruguayan conditions, a longer evaluating period is needed.