Gait analysis for patients with orthopedic joint diseases is crucial to understand their functional status. Inertial measurement unit (IMU) systems, as alternatives to optical motion capture (OMC) systems, enable gait analysis outside the laboratory. However, their accuracy requires validated before widespread clinical use. Therefore, this study evaluated the validity of the Xsens IMU system for lower-limb joint angle measurements during walking in patients with hip and knee pathology and compared the error metrics among patients with hip pathology, patients with knee pathology and healthy controls. We used OMC system and Xsens IMU systems to simultaneously collect lower-limb kinematic data of 130 patients with knee pathology, 110 patients with hip pathology and 25 healthy individuals during walking at self-selected speed. Validity was assessed using root mean square error (RMSE), amplitude difference and coefficient of determination (R²). The RMSEs for the patient groups ranged from 2.5° to 8.8°, with half of the joint angles showing acceptable accuracy and the rest tolerable accuracy. The joint angles measured by both systems were more consistent in the sagittal plane (R²:​ 0.62–0.96), with lower consistency in the frontal and transverse planes (R²:​ 0.44–0.63). Compared to the control group, both patient groups had higher RMSEs and lower R² across most joint angles. Our results suggest that the Xsens IMU system offers highly comparable sagittal plane kinematic waveforms in patients with hip and knee pathology. Caution should be taken when interpreting the frontal and transverse plane kinematics and assessing patients with severe joint deformities.