Abstract: The onset dates of the South China Sea summer monsoon(SCSSM) show a significant advancement around 1993/1994.Using sea surface temperature(SST) and reanalysis data, this interdecadal change was investigated and found to be probably attributable to the warming in the Northwest Pacific Ocean and the interdecadal differences between two types of SST anomaly in the Pacific Ocean.The results further revealed that the relationship between the timing of SCSSM onset and SST variability over the Pacific Ocean experienced a pronounced interdecadal change, accompanied by the interdecadal variation of the SCSSM onset.Eastern Pacific warming/cooling(EPW/EPC) had a primary impact on SCSSM onset in the first epoch(before 1993/1994).However, while both types of Pacific SST event affected SCSSM onset in the second epoch(after 1993/1994), the effect of central Pacific warming/cooling(CPW/CPC) was more significant.During the first epoch, EPW suppressed convective activity over the Northwest Pacific Ocean and Bay of Bengal.This suppressed convection then excited two anticyclonic circulations, located over the Philippines and the west of the Bay of Bengal, respectively.These patterns led to a delayed establishment of cross-equatorial flow, together with a significant weakening of the Bay of Bengal trough and an intensification of the western Pacific subtropical high, indicating that SCSSM onset was later than usual.The impact of EPW on the circulation variability before SCSSM onset could last from April to May.However, CPW seemingly had an insignificant influence on the preceding circulations.During the second epoch, CPW suppressed convective activity over the region from the Philippines to the eastern Bay of Bengal, which induced an anomalous anticyclonic circulation in that region.The anomalous anticyclone blocked the low trough over the Bay of Bengal and strengthened the subtropical high over the South China Sea, indicating the onset of SCSSM was earlier.The significant impact of CPW on the preceding circulations appeared in April rather than May.The impact of EPW on the circulation in April during the second epoch was similar to that during the first epoch.The anomalous wind field and convection induced by EPW, which could not sustain until May, were weaker than that forced by CPW.