Abstract. The characters of more high speed computing and much less low power dissipation are needed to settle for convolutional encodes. In this paper, we present a parallel method for convolutional encodes with SMIC 0.35µm CMOS technology; hardware design and VLSI implementation of this algorithm are also presented. Use this method, parallel circuits structure can be easily designed, which take on excellent characters of more high speed computing and low power dissipation compared with traditional serial shift register structure for convolutional encodes.