<p>Although sliding-mode control has the advantages of simple structure and easy implementation, the buffeting oscillation usually hinders its use. In this paper, we have studied the cause of the buffeting oscillation from the response property of fractional neural networks. Firstly, the feedback signal is treated as an intermittent signal owing to the periodically sampling. Secondly, the periodically feedback signal is taken as a segment function and the tie between the output and the historical input is established. Lastly, the output is separated into the transient item and the steady item. The transient is the main cause of the buffeting oscillation. Our research shows that the buffeting oscillation is relevant to the system response property, the sampling period and the control gain. Some examples verify the obtained achievements.</p>