The scarcity of modern process studies of the relationship between climatic variables and lake-water oxygen isotope (δ¹⁸O_LW) composition has restricted our understanding of the climatic significance of oxygen isotope (δ¹⁸O) records from lake sediments. We measured changes in δ¹⁸O_LW at four lakes (Nanyihu, Gonghai, Mayinghai and Xihaizi) at various locations across the East Asian summer monsoon (EASM) region, from December 2016 to January 2018. The variations in δ¹⁸O_LW at Lake Nanyihu, located in the lower reach of the Yangtze River, are controlled by precipitation oxygen isotope (δ¹⁸O_P) composition, temperature and evaporation during the non-monsoon season. However, the variations in δ¹⁸O_LW at the lakes close to the northern limit of the EASM (Gonghai, Mayinghai and Xihaizi) are affected not only by δ¹⁸O_P, temperature and evaporation, but also by the replenishment of groundwater during the non-monsoon season. In general, during the non-monsoon season, the variation of δ¹⁸O_LW exhibits a variety of patterns and the controlling factors across the EASM region are complex. During the monsoon season, although the effect of evaporation-induced isotopic enrichment still affects δ¹⁸O_LW, the variations in δ¹⁸O_LW are more influenced by the supply of abundant precipitation with lower δ¹⁸O_P to the lakes, which caused a substantial decrease in δ¹⁸O_LW across the EASM region in this season. Therefore, the variations in δ¹⁸O_LW mainly reflect the δ¹⁸O_P of the EASM during the monsoon season. However, the timing of the peak in monsoon precipitation does not correspond well to minima in δ¹⁸O_LW, and correlation analysis reveals that there is only a weak relationship between δ¹⁸O_LW and precipitation/humidity at all four lakes during the monsoon season. These findings demonstrate that although δ¹⁸O_LW can represent the δ¹⁸O_P of the EASM, it does not reflect precipitation/humidity during the monsoon season. Overall, our findings indicate that the seasonal variations of δ¹⁸O_LW across the EASM region are controlled by different climatic variables, which implies that the interpretation of lacustrine δ¹⁸O records across the EASM region as an indicator of regional effective precipitation/humidity needs to be reassessed.