BackgroundThe study was performed to qualify the source material of 4 151 pelvic radiographs for the research into the relationship between unrecognised childhood hip disorders and the development of hip osteoarthrosis, and to investigate the effect of varying degrees of pelvic tilt and rotation on the measurements of radiographic indices of hip dysplasia.Material and methods We investigated the effect of varying pelvic orientation on radiographic measurements of acetabular dysplasia using a cadaver model. Results from the cadaver study were used to validate the radiographic assessments of acetabular dysplasia in the longitudinal survey cohort of the Copenhagen City Heart Study (CCHS; Osteoarthrosis Sub-study). 1) Cadaver pelvises and proximal femurs from a male and a female donor were mounted anatomically in holding devices allowing independent inclination/reclination and rotation. An AP pelvic radiograph was recorded at each 3° increment. The most widely used radiographic parameters of hip dysplasia were assessed. 2) Critical limits of acceptable rotation and inclination/reclination of pelvises were determined on 4 151 standing, standardised pelvic radiographs of the CCHS cohort.Results Wiberg's CE angle, Sharp's angle, the xcoordinate of Goodman's Cartesian coordinate system, and the acetabular depth ratio were significantly affected by varying rotation and inclination/reclination of the cadaver pelvises. Femoral head extrusion index was not significantly affected within the applied rotation and inclination/reclination of the cadaver study. Application of the corresponding critical limits of Tönnis' foramen obturator index of 0.7-1.8 meant that 188 of