Abstract: In order to understand the temporal and spatial distribution of Coilia mystus early population in Oujiang Estuary and its influencing factors, this study used the area under the receiver operating characteristic curve (AUC) and linear regression to evaluate the predictive performance of the model. Based on the survey data of fishery resources in Oujiang Estuary, a Two-stage generalized additive model (Two-stage GAM) was used to analyze the environmental factors affecting the distribution of Coilia mystus eggs and larvae. The results showed that the main breeding period of Coilia mystus in the Oujiang Estuary was from May to July, and that temperature and salinity of surface water were the main environmental factors affecting the early population of Coilia mystus. Surface water salinity has an extremely significant effect on the distribution of eggs (P<0.01) with the suitable surface water salinity below 1. The suitable surface water temperature of eggs ranged from 24℃ to 26℃, and it has a significant effect on the distribution of larvae (P<0.05). The suitable surface water temperature of larvae ranged from 22℃ to 27℃, and the suitable surface water salinity of larvae ranged from 5 to 13. The results of cross validation showed that the prediction performance of the model for eggs was better than that of larvae. The forecast results showed that the suitable habitat for Coilia mystus eggs was relatively narrow, mainly concentrated in the waters near the upper reaches of Jiangxinyu and the lower reaches of Lingkun Island. Coilia mystus larvae mainly inhabited in the waters of Jiangxinyu to Lingkun Island. Oujiang Estuary is one of the main habitats of Coilia mystus in China, grasping the distribution characteristics of early colony of Coilia mystus in Oujiang Estuary and its relationship with environmental factors is significantly important. The results provide a scientific basis for the selection and planning of spawning ground protection area and resources management.