Abstract: Systematic errors in measurement of reflection neutrons are studied experimentally. It is found that the effects of the different factors, such as detector disturbance, detector shell, different shelves,D D neutrons, etc, are insignificant, and the repetition of the experimental system is good.