AbstractIn the quest for treatable causes of anterior knee pain, plain film skyline views of the patello-femoral joint are requested, often in 30°, 60° and 90° of flexion, to assess the functional relationships of the joint. Patello-femoral malalignment predisposes to recurrent subluxation and dislocation, articular cartilage damage and premature degenerative change. The aim of this study is to evaluate critically, by comparative assessment, the information provided by skyline views and axial computed arthro-tomography (CTA). Measurements of the patello-femoral angle (assesses patellar tilt), the congruence angle (assesses patellar lateralization) and the trochlear depth were made on Merchant's skyline views and on axial CTA on each of 50 symptomatic knees. Results are presented in graphic form with visual examples indicating a poor correlation between the two imaging methods. We conclude that skyline views are inaccurate and unsuitable primarily because they cannot be obtained in less than 30° of flexion. We suggest that skyline views have no role to play in screening for maltracking as even florid examples must be missed, and would strongly urge that no surgery be performed on their basis alone as this would result in inappropriate operations. Although computed tomography is the preferred mode of assessing patello-femoral geometry, difficulties are still encountered and the ways of circumventing them are discussed.