It is important to know the processes of precipitation in forest ecosystems due to their effects on hydrological cycle and watershed hydrology. One of these processes is throughfall. The throughfall is the portion of precipitation that falls directly to the ground or drips from the forest canopy. Although numerous studies concerning throughfall have been made in different forest ecosystems in Europe and USA, there are only five researches in Turkey. In other words, limited data are available concerning throughfall in our country. The objectives of this paper were: 1) to determine the throughfall of old oak-beach mixed forest ecosystem which is dominated in Belgrad Forest near İstanbul, and 2) to reveal relations between throughfall and precipitation in growing and dormant seasons. The study was undertaken within the Ortadere Watershed of Belgrad Forest, near İstanbul. A forest stand which represents the vegetation of watershed was selected as a study plot that comprised of old growth oak (Quercus frainetto ten.) and beech trees (Fagus orientalis Lips.), The average canopy cover was 80 %. Mean temperature is 12.3 oC and mean annual precipitation is 1129.4 mm. Mean daily temperature is 21.7 °C during August, the warmest month, and 4.2°C during January, the coldest month. In order to determine amount and variation in throughfall, two transects with 10 m. in length intersecting vertically each another were set up under forest canopy and 5 standard rain gauges were placed on the junction and endpoints of these transects. Amount of the throughfall was measured after each precipitation event. Total rainfall was recorded with a standard rain gauge placed in a clear-cut forested area nearby the study plot. Since stemflow was not measured, interception loss was not calculated. The results summarized in Table 1 show that throughfalls are 75.4 % and 82.7 % in growing and dormant seasons, respectively. Coefficients of variation of these averages are very high. This can be explained by great differences among precipitation events. The difference between throughfall of two seasons is statistically significant (P =0.004). It means that greater portion of precipitation reaches the ground in dormant season than that of growing season. Significant differences were not found among averages in sampling points of throughfall measurements. In other words, study plot has a so homogenous canopy cover that distribution of throughfall was not affected significantly. There are statistically highly significant correlations between precipitation and throughfall. Therefore, linear and curvilinear regression equations for growing and dormant seasons were developed where precipitation (mm) was the independent variable and throughfall (mm and %) was the dependent variable (fig. 3,4,5,6). Linear correlations belong to relations between precipitation in mm. and throughfall in mm. (R 2 = 0.969, Y=0.806X ; R 2 =0.987, Y=0.860X for growing and dormant season respectively) and curvilinear correlations between precipitation in mm. and throughfall in per cent (R2 = 0.523 , Y=12.75ln(X)+33.15 for growing season and R2 =0.551, Y=8.705ln(X)+51.19 for dormant season) .