Parkinson's disease (PD) is a progressive, neurodegenerative disorder, characterized by hallmark motor symptoms. Deep brain stimulation (DBS) has been used to treat advanced PD successfully. Previous studies have found that the DBS also has an effect on the electrophysiological activity of the deep brain nucleus while alleviate the PD symptoms. Here, in an attempt to gain a greater understanding of dynamic response of neural activity during subthalamic nucleus (STN) DBS for PD, local filed potentials (LFPs) were recorded from the STN during closed-loop DBS. The time frequency analysis methods short-time Fourier transform and continuous wavelet transform were used to detect the dynamic change of LFPs and the related factors which affect the length of stimulation time. The results suggest that both alpha activity and beta activity are dynamic change with electric stimulation. The delay time of DBS inhibit beta activity is about 160 ms. These results also demonstrated that the length of stimulation time are associated with the baseline amplitude, the average amplitude and the peak amplitude of beta activity. Studying the response of neural activity to electrical stimulation can reveal the electrophysiological mechanisms of DBS. Furthermore, it can improve the treatment of closed-loop DBS for PD and promote the development of intelligent neural modulation.