AbstractPurposeThe aim of this study was to measure the tibia tubercle trochlea groove distance (TT–TG) as a function of knee flexion. Our hypothesis was that there is a different pattern in healthy volunteers and patients with patella instability (PFI).MethodsThirty‐six knees of 30 patients with at least one dislocation of the patella and 30 knees of 30 healthy volunteers as control group were analysed with magnetic resonance imaging by three different observers. The TT–TG was measured in steps of 15° between 0° and 90° of knee flexion. Furthermore, the alignment of the leg (MA), the femur torsion (FTor) and the tibia torsion (TTor) was calculated.ResultsThe TT–TG was higher in patients compared to volunteers and in extension compared to flexion. This difference was statistically significant (p < 0.05). Most of the patients with a TT–TG above 20 mm in extension showed a high decrease in flexion to normal values. In some patients, this compensating mechanism fails. MA, FTor and TTor were not different in patients and control group (n.s.).ConclusionThe TT–TG distance is dynamic and decreased significantly during flexion in knees with PFI and healthy volunteers. However, there were a small number of patients in the PFI group where this compensation mechanism did not work. Therefore, the decision to perform a tibia tubercle osteotomy should not be based on one single measurement in extension or 30° of knee flexion.Level of evidenceII.