Relationship between land use and water quality varies across different river types and spatiotemporal scales. Quantitative analysis of such variations can provide scientific grounds for effective control of river water pollution. This study chose the karst river Wujiang River and the semi-karst river Jialingjiang River, to utilize the water quality monitoring data and land use data within their respective riparian buffer zones (100-500 m) and sub-watershed. Six spatial scales were considered, and redundancy analysis was employed to quantitatively explore the multi-scale impacts of land use on water quality in these different watersheds. The findings were as follows: (1) Jialing River Basin (JRB) was dominated by farmland, while forest land in the Wujiang River Basin (WRB). The aggregation and the maximum patch index of the two basins were similar. Aggregation showed no significant differences across spatial scales, and the maximum patch index decreased with the increasing spatial scale. Edge density was higher in JRB than that in WRB, and it decreases with increasing spatial scale, showing an opposite trend in WRB. The Shannon diversity index was higher in JRB, and it decreases with increasing spatial scale, while in WRB, it first increases and then decreases with spatial scale. (2) The concentration of most water quality indicators was higher in the wet season than in the dry season. Total nitrogen was a key indicator of water quality pollution in both basins, and total phosphorus was a potential nutrient pollutant. Overall, the water quality in WRB was better than that in JRB. The proportion of karst landform area and differences in land use were important reasons for the differences in water quality between the two basins. (3) Land use in JRB had the highest explanatory power for water quality variables in the dry season and at the 400 m riparian buffer zone. In WRB, land use had the highest explanatory power for water quality variables in the wet season and at the sub-watershed scale. Generally, land use in WRB contributed more to water quality changes, comparing to JRB. (4) Different land use indices had varying impacts on water quality. Construction land, edge density, farmland, and water bodies significantly impacted the JRB water quality, with water bodies exerting a positive effect. Water bodies, the Shannon diversity index, and farmland were the key indices that affected the WRB water quality, with farmland having a certain positive impact. Therefore, different types of rivers should, within the optimal spatial scale, plan the area of farmland and construction land based on the landform characteristics of the basin, increase the forest coverage rate in riparian zones, and promote ecological fertilization and ecological control technologies to enhance river water environmental safety and sustainable water resource utilization.