To know breast doses for 2D digital mammography and 3D tomosynthesis equipment ensure that determining risks for breast which organ is radiosensitive. In this study, main objective is to construct a mathematical model, using multiple linear regression analysis method, which takes into account the exam specific parameters such as patient age, breast thickness, breast surface area, force, tube voltage and effective tube current and provides an estimation of the actual breast dose during 2D digital mammography and 3D tomosynthesis exams.A dosimetric quality control test has been carried out to determine the reliabilities of Entrance Skin Dose (ESD) and Average Glandular Dose (AGD) values given by the digital mammography device following each exam. Patient data have been derived from the Picture Archiving and Communication System (PACS) for a retrospective evaluation. Radiologists have separated patients into 4 groups represented as A, B, C and D based on the BI-RADS criteria which account for breast composition. Each group included 50 patients to yield statistically meaningful results. A database has been constructed that contains dose information and the exposure parameters specific to mammography exams. Breast surface area determined by manually drawn regions of interest (ROI) has also been included in the database. AGD and ESD values have separately been analyzed with respect to each exposure parameter. It has been investigated whether there were significant differences between the data sets by applying Kruskal–Wallis analysis. Mathematical model has been constructed on the base of multiple regression analysis to determine the effects of independent variables on AGD and ESD.The dosimetric quality control of the digital mammography device was performed and it was found that the device was operating under optimum conditions. Kruskal Wallis tests carried out between AGD, ESD and independent variables have resulted in p values lower than 0.05. In the model based on multiple regression, AGD and ESD corresponded to the 95.4% and 93.9% of variance, respectively.The constructed model provides a practical method to calculate AGD and ESD by using the exposure parameters without need to any conversion factors.