Pelvic incidence (PI) is often quantified in patients undergoing total hip arthroplasty. Errors in radiographic PI measurements can affect clinical outcomes. The purposes of this study were (1) to evaluate the error in radiographic PI measurement in patients with hip osteoarthritis (OA) and (2) to analyze the factors related to the error. Radiographs and computer tomography (CT) images of 100 patients (24 men and 76 women;​ mean age 63.7 years) with unilateral OA were reviewed. The error in radiographic PI measurement was defined as the difference between the radiographic measurement of the PI (rPI) and the accurate value of PI measured using CT images (cPI). Factors related to the error in the rPI were analyzed, including the coronal and axial rotation of the pelvis on lateral radiographs. The degree of coronal and axial rotation was expressed as the angle of rotation around the anteroposterior and craniocaudal axes. The mean rPI was significantly larger than the cPI (57.8° and 54.1°, p < 0.01). The error in the rPI was 3.6° on average and 15.8° at maximum. The mean coronal and axial rotation of the pelvis was 9.6° and 4.4°, respectively. The error in the rPI positively correlated with coronal pelvic rotation and rPI, and negatively correlated with axial pelvic rotation (p < 0.01, r = 0.35, 0.43, and −0.45, respectively). The rPI was 3.6° larger on average than the cPI in patients with hip OA. Coronal and axial rotation of the pelvis and a large PI were related to the error in the rPI.