PurposePatellar dislocation is associated with a range of anatomical abnormalities affecting the trochlea, extensor mechanism and the tibia. The relationship between patellofemoral instability and rotational abnormalities of the posterior condyles, trochlear groove and proximal tibia has not been adequately determined. This study aimed to identify the frequency and severity of anatomical risk factors to determine their relative contribution to patellofemoral instability. MethodsA retrospective morphological study was undertaken comparing multiple anatomical measurements with magnetic resonance imaging of 50 patients with patellofemoral instability to an age‐ and gender‐matched Control group (n = 50). Several techniques were assessed measuring both femoral and tibial axial asymmetry. A new measurement, tibial rotational asymmetry, comparing a line between the midpoints of the collateral ligaments to the axis between the patellar tendon and posterior cruciate ligament, was assessed for its association with patellofemoral instability. ResultsCompared to the controls, the patellofemoral instability group demonstrated a significant difference in tibial rotational asymmetry, with a mean of 2.9° (SD 3.2°) externally rotated vs − 1.6° (SD 2.2°) in the control group. Significant differences were also demonstrated regarding the sulcus angle, tibial tubercle–trochlear groove distance, tibial tubercle–posterior cruciate ligament distance, patellar size and the Insall–Salvati ratio. There were no differences between groups regarding the lengths of the posterior condyles, the heights of the trochlear ridges or lateralisation of the trochlear groove. Further analysis of the patellofemoral instability group revealed a subgroup of males with normal anatomy (7/50) and a subgroup of females with isolated patella alta (7/50). ConclusionPatellofemoral instability is associated with tibial rotational asymmetry due to lateralisation of the tibial tubercle. It is also associated with patella alta and reduced trochlear groove depth. The femoral axial shape is otherwise unchanged. Level of evidenceIII.