BACKGROUND: Although studies have provided the estimates of floods-diarrhoea associations, little is known about the lag effect, effect modification, and attributable risk. Based on Sichuan, China, an uneven socio-economic development province with plateau, basin, and mountain terrains spanning different climatic zones, we aimed to systematically examine the impacts of floods on diarrheal morbidity. METHODS: We retrieved information on daily diarrheal cases, floods, meteorological variables, and annual socio-economic characteristics for 21 cities in Sichuan from January 1, 2017 to December 31, 2019. We fitted time-series Poisson models to estimate the city-specific floods-diarrhoea relation over the lags of 0-14 days, and then pooled them using meta-analysis for cumulative and lag effects. We further employed meta-regression to explore potential effect modifiers and identify effect modification. We calculated the attributable diarrheal cases and fraction of attributable morbidity within the framework of the distributed lag model. RESULTS: Floods had a significant cumulative association with diarrhoea at the provincial level, but varied by regions and cities. The effects of the floods appeared on the second day after the floods and lasted for 5 days. Floods-diarrhoea relations were modified by three effect modifiers, with stronger flood effects on diarrhoea found in areas with higher air pressure, lower diurnal temperature range, or warmer temperature. Floods were responsible for advancing a fraction of diarrhoea, corresponding to 0.25% within the study period and 0.48% within the flood season. CONCLUSIONS: The impacts imposed by floods were mainly distributed within the first week. The floods-diarrhoea relations varied by geographic and climatic conditions. The diarrheal burden attributable to floods is currently low in Sichuan, but this figure could increase with the exposure more intensive and the effect modifiers more detrimental in the future. Our findings are expected to provide evidence for the formulation of temporal- and spatial-specific strategies to reduce potential risks of flood-related diarrhoea.