In this paper, wavelet transform was applied in the sea-level change multi-scale cycle analysis of monthly average data signal in East China Sea, and future sea-level change was predicted by the way of winters exponential smoothing. The results show: 1) sea level showed upward trend in volatility in East China between 1992 to 2009; 2) sea level data denoising signal showed upward trend more obvious than the original signal. Compared to annual maximum of sea level height(in annual summer and autumn), sea level rise was more evident, and the average growth rate is up to about 0.78 cm/a, and it was 2 times of the average annual growth rate; 3) future sea level changes (after 2006) will maintain an upward trend with fluctuations in East China Sea. Annual average growth rate will be about 0.49 cm/a. By 2015, sea level rise rate will be about 4 cm to 5 cm higher than 2006. The rate of sea level rise will increase to about 14 cm to 15 cm by the year 2030 on the basis of the year 2006.