Abstract: The interdecadal and interannual components of global sea surface temperature anomaly (SSTA) were obtained by periodical analysis, and their significant variations over different areas of global sea and seasonal change were displayed by variance analysis. Then using singular value decomposition (SVD), the simultaneous and time-lag correlations of interdecadal sea surface anomaly and winter temperature/summer rainfall of China were investigated. Results show that the interdecadal SSTA is more evident than the interannual one both in winter and summer, and positive SSTA is usually assotiated with positive winter temperature anomaly in North China and positive rainfall anomaly in the Changjiang valley and South China, but negative rainfall anomaly in North China, and vice versa.