Background <p>Surface water is a major source of domestic water supply in China, yet its quality remains uneven despite recent improvements. Chronic exposure to surface-water pollution has been implicated in liver injury, but evidence linking water quality to cirrhosis mortality—particularly accounting for delayed and non-linear effects—remains limited. We investigated the spatiotemporal associations between surface water quality and cirrhosis mortality in China.</p> Methods <p>We conducted a nationwide ecological study integrating surface water monitoring data and cirrhosis mortality data from 2013 to 2021. Distributed lag non-linear models (DLNMs) were applied to assess non-linear and lagged associations (0-2 years) between overall water-quality classes and key pollution indicators, including total phosphorus (TP), chemical oxygen demand, dichromate method (COD_Cr), and five-day biochemical oxygen demand (BOD₅). Analyses were stratified by sex and urban-rural residence.</p> Results <p>Surface water quality was associated with cirrhosis mortality in a non-linear, delayed, and population-specific manner. Moderate and poor water quality exhibited distinct temporal risk patterns, with sustained exposure to moderate pollution associated with a higher cumulative mortality risk. TP and COD_Cr showed clear non-linear associations with mortality that intensified at longer lags, whereas BOD₅ displayed a U-shaped relationship with relative risks remaining below unity. Stronger and more persistent effects were observed among males and rural residents.</p> Conclusions <p>Our findings indicate that long-term exposure to moderate surface-water pollution and persistent nutrient and organic contamination contribute to cirrhosis mortality in China. Water-quality management strategies should account for delayed health effects and population vulnerability, with targeted interventions prioritising sustained pollution reduction and protection of high-risk groups.</p> Graphical Abstract <p></p>