Abstract: The characteristics of the river basin landscape have a significant impact on the migration and output of nitrogen (N) and phosphorus (P) in river basins, but its relationship with the output load of river N and P and its contribution need to be further studied. This study takes the Jinjing River Basin as the research area, and quantitatively estimates the river N and P output load based on long~term hydrology and water quality positioning observation data. Based on the classification of landscape index combined with the core landscape index of different landscape scales selected by traditional statistical analysis, redundancy analysis (RDA) and variance analysis (VPA) are used to clarify that the characteristics of landscape pattern at different scales in the typical subtropical agro~forestry mixed watershed are important for river N, P The influence of output load. The results show that the output load of N and P in the river basin during the flood period and the dry and dry period is significantly different in time and space. The output load of TN, NH₄⁺-N, NO₃^--N, TP, and DIP in the river basin during the flood period are 2.61-9.03, 0.34-5.34, and 0.34-5.34, respectively. 1.26-5.03, 0.15-0.50, 0.09-0.23 kg/(hm2.a); the flat and dry seasons were 1.55-6.68, 0.29-2.91, 0.74-2.91, 0.10-0.46, and 0.05-0.20 kg/(hm2.a), respectively. Different landscape scale levels and land use types, landscape characteristics have different effects on N and P output loads, but at the same scale level, the impact of various types of landscape indexes on the N and P output loads of rivers during high and low water periods is basically the same . At the landscape scale, the increase in the concentration of different types of patches and the increase in landscape diversity will lead to an increase in the output load of river N and P; the output load of river N and P is related to the large area, high concentration, and high connectivity of farmland and residential land. Blocks are significantly positively correlated, and significantly negatively correlated with high~aggregation and large~area woodland patches. The main impact index categories of farmland patches on river N and P loads in wet and dry periods are area~margin and aggregation index respectively. The impacts of forest land and residential land patches at different water level periods remained unchanged, and they were respectively the area~margin and agglomeration category indexes. The results of this study can provide theoretical support and scientific basis for the management of agricultural non~point source pollution in the river basin from the perspective of landscape planning.