With the enhancement of sewage and drainage system infrastructure and the increase in water quality purification plant effluent standards in urban regions of China, point-source pollution in most urban areas has been effectively controlled. However, due to the rapid population growth and expansion of urban areas during the urbanization process, non-point source pollution plays an increasingly important role in the pollution of urban surface water. This study addresses the existing challenges in accurately quantifying loads and identifying characteristics of non-point source pollution in high density urban areas by taking the Guanlan River basin as a case study area. The Soil and Water Assessment Tool (SWAT) model was used to simulate the total nitrogen (TN) and total phosphorus (TP) loads originating from non-point source pollution in the basin, as well as to identify their spatiotemporal patterns. Through establishing two pollution mitigation measures comprising six scenarios, in conjunction with the SWAT model, the mitigated efficiency of TN and TP levels for each scenario in the case study area in 2022 was evaluated. The simulated values of the SWAT model revealed a good agreement with the observed data for runoff, TN, and TP. The SWAT model was employed to examine spatiotemporal characteristics of non-point source pollution TN and TP loads in the study area in 2022. Spatially, some sub-basins exhibited relatively high levels, with TN levels varying between 1036.1 and 1523.9 kg/km2 and TP levels ranging from 473.5 to 572.9 kg/km2. Temporally, the most severe precipitation runoff pollution was recorded in May, where TN and TP loads reached 252.47 kg/km2 a nd 215.91 k g/km2, respectively. It was observed that both filter strips and grassed waterways measures have been proven effective in mitigating non-point source pollution in the case study area. These engineering measures indicate the appealing potential in reducing the TN and TP from non-point source pollution in high density urban areas China-wide.