This review article presents several pitfalls and limitations of image interpretation of anatomical risk factors for patellar instability. The most important imaging examinations for the work-up of patients with patellar instability are the true lateral radiograph and transverse computed tomography (CT) or MR images of the knee. Primary anatomical risk factors are an insufficient medial patellofemoral ligament (MPFL), patella alta, trochlear dysplasia, increased distance from the tibial tuberosity to the trochlear groove (TTTG), and torsional limb parameters. Limitations of the Caton-Deschamps index are related to the clear identification of the patellar and tibial articular margin. Classification of trochlear dysplasia according to the Dejour system on radiographs and MR images revealed a weak reliability. The comparability of TTTG values obtained on CT and MR images at various flexion angles and different varus alignments of the knee is limited. Thus MRI performed with a dedicated knee coil may underestimate the TTTG distance compared with CT images. Increased lateral patellar tilt is a consequence of primary anatomical risk factors rather than an independent anatomical risk factor for patellar instability. The pretest likelihood of a torn MPFL on MR images is very high after an acute episode of lateral patellar dislocation. Surgical restoration of the patellofemoral joint stability addresses the complex multifactorial biomechanics by a custom-made management such as MPFL reconstruction, sulcus-deepening trochleoplasty, as well as medialization and distalization of the tibial tubercle. Quantification of anatomical risk factors for patellar instability in each person is important for highly individual treatment.