The development of more effective fixation devices for reapproximating and immobilizing the sternum after open-heart surgery is limited by current methods for evaluating these devices. In particular, precise emulation of in vivo sternal loading has not been achieved in controlled model systems. The present study is an initial effort to determine the in vivo loading parameters needed to improve current in vitro and in silico (computational) models. Towards this goal, the direction, magnitude, and distribution of loading along a midline sternotomy were characterized in a porcine model. Two instrumented plating systems were used to measure the forces across the bisected sternum in four anaesthetized Yorkshire pigs during spontaneous breathing, ventilated breathing, and coughing for four treatments: live, cadaveric, embalmed, and refrigerated. Changes in forces incurred by death and embalming were also investigated to evaluate the potential applicability of cadavers as models for testing sternal fixation devices. The magnitudes of the respiratory forces in three orthogonal directions ranged from , many fold smaller than previously estimated. Dynamic forces were highest in the lateral direction during coughing and low in all directions during normal breathing. No significant differences in force were found between the four treatments, most likely due to the unexpectedly low magnitude of forces in all groups. These results provide the first measurements of in vivo sternal forces and indicate that small cyclic fatigue loads rather than large quasistatic loads should be applied in future model systems to best evaluate the mechanical performance of fixation devices.