PurposeThe aim of this study was to identify the most accurate and reliable quantitative radiographic parameters for assessing vertical and horizontal instability in different Rockwood grades of acromioclavicular joint (ACJ) separations. Furthermore, the effect of projectional variation on these parameters was investigated in obtaining lateral Alexander view radiographs.MethodsA Sawbone model of a scapula with clavicle was mounted on a holding device, and acromioclavicular dislocations as per the Rockwood classification system were simulated with the addition of horizontal posterior displacement. Projectional variations for each injury type were performed by tilting/rotating the Sawbone construct in the coronal, sagittal or axial plane. Radiographic imaging in the form of an anterior–posterior Zanca view and a lateral Alexander view were taken for each injury type and each projectional variation. Five newly defined radiographic parameters for assessing horizontal and vertical displacement as well as commonly used coracoclavicular distance view were measured. Reliability, validity and the effect of projectional variation were investigated for these radiographic measurements.ResultsAll radiographic parameters showed excellent intra- and interobserver reliability. The validity was excellent for the acromial centre line to dorsal clavicle (AC–DC) in vertical displacement and for the glenoid centre line to posterior clavicle (GC–PC) in horizontal displacement, whilst the remaining measurements showed moderate validity. For AC–DC and GC–PC, convergent validity expressed strong correlation to the effective distance and discriminant validity demonstrated its ability to differentiate between various grades of ACJ dislocations. The effect of projectional variation increased with the degree of deviation and was maximal (3 mm) for AC–DC in 20° anteverted malpositioning and for GC–PC in 20° retroverted malpositioning.ConclusionsAC–DC and the GC–PC are two novel quantitative radiographic parameters of vertical and horizontal instability in ACJ dislocations that demonstrate excellent reliability and validity with reasonable inertness to malpositioning. The use of AC–DC for assessing vertical displacement and GC–PC for assessing horizontal displacement in a single Alexander view is recommended to guide the appropriate management of ACJ dislocations. A better appreciation of the degree of horizontal instability, especially in lower Rockwood grades (II, III) of ACJ dislocations, may improve management of these controversial injuries.