Based on analysis of observations, we defined the starting and ending time and the total rainfall of the spring persistent rainfall in Jiangnan of China (SPRJ) from 1982 to 2014, and investigated the climatic characters of the SPRJ as well as its interannual variation. The relationship between the SST anomaly of Nino3.4 region in the preceding winter and the SPRJ and its physical mechanism were further studied. Results show that there is a significant positive correlation between the preceding winter Nino3.4 index and total SPRJ. The warm water can trigger an anomalous Walker circulation that leads to significant abnormal descending motions and corresponding low-level anticyclonic circulation near 120°E at the equator. The strengthened southwesterly winds of the anticyclone in the low level of the South China Sea facilitate more water vapor transport from the South China Sea to Jiangnan of China. On the other hand, the Indian Ocean SST anomaly associated with El Niño events can induce abnormal low level easterly winds in tropical Indian Ocean and anticyclone in the north of the Bay of Bengal, which also promotes the water vapor transportation. Meanwhile, the westerly anomalies in the upper troposphere above South Asia enhances divergence and pumping above Jiangnan of China, and thus are favorable for ascending motions and more SPRJ. In contrast, there is less SPRJ following a La Niña event. Moreover, the influence of El Niño on the SPRJ changes with its original intensity. With a strong El Niño in the preceding winter, the SSTA in the Pacific can persist to the following spring and there will be more SPRJ rainfall; with a weak El Niño in the preceding winter, however, the SSTA in the Pacific cannot persist to the following spring and the SPRJ total rainfall will decrease. Besides, when considering the combined effects of Nino3.4 index and Antarctic Oscillation Index and Indian Ocean Basin Mode of the preceding winter, the seasonal prediction is improved. Thus the multiple linear regression of the three predictors is useful for the prediction of the SPRJ.