Background: Proprioception can be assessed by measuring joint position sense (JPS). Most studies have focused on JPS of the knee joint while literature for other joints especially for hip JPS is scarce. Although some studies have evaluated proprioception of the knee joint, the reproducibility of methods has rarely been investigated. Aims: To estimate intrasession reliability and agreement of an active–active JPS test for hip flexion/abduction and knee flexion in healthy older adults. Methods: Nineteen healthy older adults participated in this study. The proprioception of the hip (flexion and abduction) and knee (flexion) were assessed in both legs using the “active–active” reproduction technique. Intraclass correlation coefficient (ICC), standard error of measurement (SEM), and limits of agreement (LOA) were estimated for relative angular error (RE), absolute angular error (AE), and variable angular error (VE). Results: Reliability of our JPS test was substantial to almost perfect for the RE for both joints and legs (ICC values ranging from 0.75 to 0.93). We also found that the ICC values for AE were substantial for knee flexion and hip abduction of the left and right leg. The ICC results of VE showed poor reliability for hip and knee joints. SEM and LOA values for hip abduction were generally lower than for hip and knee flexion, indicating lower measurement error or more precise scores for the proprioception test of hip abduction. Conclusion(s): Proprioceptive acuity of the knee and hip joints in healthy older adults can be reliably assessed with an active–active procedure in a standing position with respect to relative and absolute error.