In order to improve the drawback of the high peak-to-average power ratio (PAPR) on the multi-carrier systems, in this paper we derive a statistical characterization approach of the time domain signal power variance metric by means of every user¡¦s spreading code combination and subcarrier scrambling. We obviously reduce the PAPR on the multi-carrier code division multiple access (MC-CDMA) systems by resorting the combination of spreading codes and scrambling the polarities of subcarriers. Due to the large calculative number of exhaustive search, we use a low complexity Replacement Search Method (RSM) to reduce the calculation times of searching, and obtain a good result. Moreover, we can get a better PAPR reduction performance by increasing the number of iteration.