The trains of 200 ms biphasic square pulses with the width of 9 ms delivered at 50 Hz were found to be the most suitable and effective mean as stimulation in FES system of restoring the blink function in unilateral facial nerve paralysis rabbit model. FES system is a reliable tool for these patients. Facial paralysis affects thousands of people every year. Many will have long term facial difficulties and the loss of the ability to blink the eye, which can lead to potential loss of the eye. Although many treatments exist, no one approach corrects all the deficits associated with the loss of orbicularis oculi function. FES is a means of providing movement in paralysed muscles to assist with practical activities and one possible way of restoring blink and other functions in these patients. Although some previous researches had investigated the effect of simple FES system on restoration of paralyzed facial muscles, there is still controversy about the appropriate details of the most effective stimulating pulses, such as the frequency, wave pattern and pulse width. Our aim is to find out the parameters of the most appropriate and effective stimulatin verify it by a simple FES system. 24 healthy adult male New Zealand white rabbits were accepted the surgery of right side facial nerve main trunk transaction under general anesthesia as the unilateral facial nerve paralysis models. The platinum tungsten alloy electrodes were implanted in orbicularis oculi muscle. The parameters of stimulus pulses were set to a 200 ms biphasic pulse with different waveforms (square, sine and triangle), different frequencies (25, 50, 100 Hz) and different widths from 1 to 9 ms. Next, we set up a simple FES system to verify the previous results as the stimulus signal. We observed the movement of the both sides of eyelid when eye blink induced by different kinds of pulses. In all animals, the three kinds of waveforms pulse with frequency of 25 Hz could not evoke the smooth blink movement. But the pulses with frequency of 50 and 100 Hz can achieve this effect. The voltage threshold of the square pulse was lower than that of the sine pulse and triangle pulse. With the increase of pulse width from 1 to 9 ms, the voltage threshold decreased gradually. The voltage threshold of the pulse with frequency of 100 Hz was obviously lower than that of 50 Hz. But the amount of total charge of the stimulation pulse of 100 Hz was significantly more than that of 50 Hz. In addition, when the FES system was turned on, the eye blink on the affected side with the stimulation pulses that were set by the previous step results was successfully aroused by the blink movement as a trigger on the contralateral.