The outcome of patellofemoraljoint reconstructive surgery has been found to be variable. This study presents patient-specific models of the patellofemoraljoints of 13 patients diagnosed with osteoarthritis. The model is a mathematical model using quasi-static equilibrium analysis and the model inputs are patient data is acquired from a combination of two series of magnetic resonance images. Four tibial tuberosity transfer surgeries are simulated for each patient, namely: 15 mm anterior shift, 20 mm anterior shift, 8 mm anterior along with 8 mm medial shift, and 15 mm anterior along with 8 mm medial shift. The simulated surgeries produced a statistically significant decrease in mean contact stress relative to pre-surgical conditions (p=0.004), though no statistical difference was found between the different procedures. No one procedure consistently demonstrated the best outcome for all patients. In fact, certain surgical procedures increased peak stresses in some subjects while decreasing it in others. With the demonstrated variability of outcome, the utility of the model presented for surgical planning becomes apparent.