Assessment of knee joint position sense by the method of reproduction of passive position within the same limb was investigated in 13 normal subjects. It was found that the reliability of this measurement is not good under either active or passive reproductions. The mean error in this active trials is higher than that in the passive trials and that is more of a tendency to overshoot during active trials. People perceiving passive and active conditions seem to involve different mechanism. This different peripheral feedback may alter the central perception of position. In future assessments the trial position. and the testing condition may need to be similar to test both active and passive mechanisms contributing to position sense.