<p>Integrated water quality indices are widely used to summarise complex monitoring datasets and support long-term river assessments. However, their ability to capture structural water quality changes over extended periods, particularly in large rivers characterised by strong hydrological mixing, remains insufficiently evaluated. This study uses long-term monitoring data from a large river system and its major tributaries to assess how integrated indices reflect long-term nutrient dynamics. The analysis is based on annual mean concentrations of key physicochemical parameters for 2001–2024. A weighted arithmetic Water Quality Index (WQI) was used to evaluate temporal and spatial patterns, long-term trends were assessed using non-parametric methods, and principal component analysis (PCA) was applied to examine both the internal structure of the WQI and underlying hydrochemical variability. Results show that WQI values in the main river channel remained relatively stable and were predominantly classified as good, despite statistically significant changes in individual nutrient parameters. In contrast, several tributaries exhibited greater variability and poorer water quality during the early monitoring period. PCA of WQI sub-indices revealed strong dominance of the phosphorus component, which explained nearly all long-term variability in the aggregated index, while PCA of physicochemical data identified nutrient enrichment and organic pollution as the main gradients shaping water quality variability. These findings indicate that in large rivers, integrated water quality indices may mask structurally important nutrient dynamics over long time scales, highlighting the need to complement integrated indices with multivariate analyses when interpreting long-term water quality change.</p>