By investigating the contents of five heavy metals Cu, Zn, Pb, Hg, and As in the water body of the Yangtze River Estuary area from 2004 to 2017, the change trends and influencing factors of water heavy metal contents in the three areas of Beizhi, Nanzhi Beigang and Beigang Beisha were analyzed and the contents of heavy metals were predicted in the next few years. The results show that Cu and Pb showed a downward trend as a whole from 2011 to 2017, and the decline rate was lower than that from 2004 to 2008. Zn and As were in a stable downward trend from 2011 to 2017. The downward trend of Hg is small, and five heavy metals increased in 2011 and 2015 when pollutant emissions increased, and regions with larger runoffs experienced larger increases; Analyzing the influencing factors, it is concluded that due to the high content of heavy metals in the early water body, the effectiveness of the treatment investment is greater than the impact of pollutant discharge. Therefore, the contents of Cu, Zn, Pb, and As from 2004 to 2008 were significantly negatively correlated with the environmental treatment investment, and the correlation with the amount of discharge is not significant; After the heavy metal content was controlled to a low level, the difficulty of governance increased and the effectiveness was weakened. Pollutant emissions became the main influencing factor for the control of heavy metal content. Therefore, the correlation between heavy metal content and environmental governance investment from 2011 to 2017 was not significant, and there is a significant positive correlation with the amount of discharge; Due to the differences in salinity and runoff in the three regions of the Yangtze River Estuary, ARIMA model was used to predict the heavy metals in different regions, and the heavy metal content in the water bodies of the three regions at the mouth of the Yangtze River Estuary from 2020 to 2022 was predicted to be low and the change trend is relatively stable, the model has a high accuracy, the error is between 5.19% and 11.82%. The regional prediction can highlight the heavy metal content and change characteristics of different regions, the prediction results are more targeted, and can be used for future governance .