Estuaries are generally considered to be important sources of atmospheric CO2. However, the differences between estuaries, and inadequate observations of partial pressure of CO2 in estuarine water (pCO2water) hamper global estuarine CO2 budgeting. In this study, the longitudinal distribution of CO2 in the waters of Modaomen (MSE) and Lingdingyang (LSE), two sub-estuaries of the Pearl River Estuary (PRE), and its influencing mechanism are studied. The change in the distribution of pCO2water along the distance from the upstream estuary to the ocean between LSE and MSE was significantly different. pCO2water at the LSE ranges from 238 to 7267 µatm, whereas the MSE ranges from 406 to 3078 µatm. Stronger microbial respiration and relatively long water retention times were the main influences that led to higher pCO2water at LSE than at MSE. Seasonally, the increase of soil CO2 into the water in the upstream basin caused by precipitation is the potential influencing factor that the water pCO2water in the flood season is higher than in the dry season. PRE was a net source of atmospheric CO2 with an average annual water-air flux of 41.2 ± 33.3mmolm-2day-1. Our results suggest that the differences in longitudinal gradients of pCO2water between estuaries in the same region and the effects of different gas transport velocity models on CO2 emission estimates need to be considered in estuarine CO2 emission budgeting.