Detailed information about apple phenology allows a grower to manage the orchard efficiently and to decide how many fruits per tree harvest. In general, eight stages have been identified for apple development starting with green tip until petal fall. Mathematical models can be developed to predict crop phenological stages. The objective of this study was to model apple phenology as a function of the physiological time using cumulative degree days (GDD). The cultivars 'Cripps Pink', 'Gala', and 'Red Delicious' were monitored twice weekly during four growing seasons for eleven different locations in the main apple growing regions of eastern Washington. Thirty buds were selected and the number of buds at a particular stage was registered. Daily temperature data came from the closest automated weather station of AgWeatherNet to the orchard where the samples were collected. The field observations showed that the stages were not mutually exclusive with at least two stages overlapping during apple development. The approach for the model was based on the duration or accumulation of GDDs for each individual stage from the first to the last observation of each stage. To determine GDD, the base temperature, i.e. 6.11°C was subtracted from the average temperature. For each stage, the lower and upper limits for the duration were calculated as the average of the minimum and the maximum GDDs for each stage. A comparison of the stage distribution among seasons, cultivars, locations and their interactions was conducted using GLIMMIX SAS procedure. There was significant effect of the factors and their interactions, indicating the complexity of apple development. The duration of each phenological stage was similar for all of the cultivars however cultivar 'Gala' started later than the other two cultivars for each stage. The model performance was good. A prototype of the model will be implemented on the AgWeatherNet portal (www.weather.wsu.edu).