Radiographic measurements of femoral varus are used to determine if intervention to correct femoral deformity is required, and to calculate the required correction. The varus angle is defined as the angle between the proximal femoral long axis (PFLA) and an axis tangential to the distal femoral condyles. Ideal positioning of the femur in terms of zero elevation relative to the radiographic plane or zero rotation along the long axis of the femur may be difficult to obtain in practice. The clinician may thus be tempted to determine varus values from less than ideal radiographs. Fourteen isolated right femora were positioned at zero elevation (supracondylar eminence at same distance from the radiographic plate as the lesser trochanter, as defined in previous studies) and with the caudal condyles in contact with the radiography cassette (0° rotation) and subsequently rotated internally and externally by 5° and 10° using plastic wedges. Accuracy of rotation was within +1°. Digital radiographs were obtained at each position. Varus angles were measured using ImageJ, employing two definitions of PFLA. Mean varus angles increased with 10° of either internal or external rotation with both PFLA definitions, but at 5° increased slightly with one definition and decreased with the other. Scatter plots indicated that not all femora exhibited the same trend in change in varus angle. Significant differences (p<0.05) were observed between varus measurements for the different PFLA definitions and between rotation angles. The effect of rotation on varus angle measurements in these femoral specimens contradicts a previous report using CT. The most probable explanation is the difference in femoral positioning: the CT study used a slightly elevated position compared to that in this study, resulting in better visualisation of the condylar articular surfaces. Zero elevation frequently results in the articular surfaces being obscured by the trochlear ridges, even when positioned perfectly. Varus measurements vary unpredictably as a result of femoral rotation and this effect is likely further influenced by femoral elevation.