Biomechanical breast modeling using finite element (FE) analysis to predict 3-D breast deformations is of interest for various biomedical applications. Currently no consensus of reliable magnitudes of mechanical breast tissue properties exists. We therefore applied 12 material properties proposed in the literature to FE simulation models derived from prone MRI breast datasets of 18 female volunteers. A gravity free starting position is computed with an iterative FE algorithm followed by the calculation of the upright position of the breast and then compared to the real breast geometry in standing position using corresponding 3-D surface scans to determine the accuracy of the simulation. Hyper-elastic constitutive models showed superior performance than linear elastic models which cannot exceed the linear Hookean domain. Within the group of applied hyper-elastic material models those proposed by Tanner et al. (Med Phys 33:1758-1769, 2006) and Rajagopal et al. (Acad Radiol 15:1425-1436, 2008) performed significantly (p < 0.01) better than other material models. The advantage of the method presented is its non-invasive character by combining 3-D volume and surface imaging with automated FE analysis. Thus, reliable biomechanical breast models based on the presented methods can be applied in future to derive patient-specific material parameter sets to improve a wide range of healthcare applications.