Abstract: Aimed at the problems of the parameter estimation of non-periodic long-code Direct Sequence Spread Spectrum (DSSS) , in this paper, a representation model for non-periodic long-code DSSS is built up, and the autocorrelation function of its delay-product is derived and analyzed. The autocorrelation function has periodic peaks, and a reciprocal of the space between two adjoining peaks is the chip-rate of pseudo-random sequence. The effect of the noise to the autocorrelation function is also analyzed, and theoretically the effect of the noise could be canceled by choosing the right delay . On the basis of these properties an estimation method of the chip rate of pseudo-random sequence of non-periodic long-code DSSS is proposed. At first, the autocorrelation is calculated after the received base signal is delayed and multiplied. Then the peak which is generated by noise is canceled. At last, the interval between two adjoining peaks is estimated and the chip rate is got. The computer simulations results show that the proposed method is effective.