The objective of this paper was to estimate the dormant season interception of an oak-beech mixed forest with the growing season interception data by using an artificial neural network model. Precipitation and throughfall data were used to find out the amount of interception amounts in growing and dormant seasons. There is statistical difference (P =0,004) between two seasons in terms of interception.  The data was divided into two groups of testing and estimating and was then integrated to the articial neural network model. Growing season precipitation data (total precipitation-throughfall) were used as inputs while interception data were used as outputs.. Dormant season precipitation (total precipitationthroughfall) were used as estimating group inputs data. A performance evaluation composed of regression and mean squared error, was performed between interception values of dormant season and estimating interception values with artificial neural network. Significant linear correlation was found between estimated and measured interception values with a high determination coefficiant (R2 = 0.90 for dormant season respectively) and low mean square error (MSE = 3.47) . Models presented in this study are applicable to stands of similar features. In other words, if throughfall and total precipitation values are known, these models provide the researchers with the opportunity to estimate interception amounts of different seasons in ecosystems of stand which have similar characteristics.