BACKGROUND The efficiency and success of computer-assisted fluoroscopic navigation systems mainly depend on the quality of the process of image acquisition: obtaining the correct view of anatomic structures, relative orientation of multiplanar X-ray images and the necessary amount of radiation dose. These systems may be optimised by using a system called zero-dose c-arm navigation (ZDCAN). We investigate whether the available computer-assisted systems may be used to navigate the c-arm before image acquisition to obtain X-ray images with maximised accuracy and minimal radiation exposure. METHODS Based on position data of an optical tracking system combined with statistical deformable bone models, ZDCAN is able to generate a real-time preview of expected X-ray images of the lower extremities without using radiation. We performed a cadaver study on six full-body specimens comparing the zero-dose approach to conventional positioning of the c-arm in order to evaluate efficiency and accuracy. Eight users acquired two perpendicular X-ray images of the hip, the knee and the femoral diaphysis. RESULTS The number of X-ray images required to get a satisfying picture could be reduced to seven using the zero-dose approach; the conventional approach needed 11 images. The mean time did not differ significantly. Regarding the image quality, using ZDCAN quasi-orthogonality could be reached while the conventional approach showed a large variation of the relative orientation. CONCLUSION Using ZDCAN, the amount of radiation can be reduced by requiring less X-ray images as well as reaching better accuracy.