Urbanization has significant effects to ecosystems in watersheds, but the link between surface water quality and the dissolved organic matter (DOM) composition is poorly understood. We investigated the fluorescent intensities (F_max) of DOM components and examined their correlation with the water quality parameters in peri-urban (Zhangxi River) and urban (Lu River) watersheds in Ningbo, East China. DOM quality was measured by fluorescent excitation-emission matrices (EEMs) coupled with parallel factor analysis (PARAFAC). Terrestrial humic-like components (C1 and C2) and protein-like component (C3) were derived by the PARAFAC model. We found more serious water pollution (significantly higher values for most water quality parameters, i.e. chemical oxygen demand (COD), chlorophyll a (Chl-a), dissolved organic carbon (DOC), ammonia and total suspended solids (TSS)) in urban than peri-urban watersheds. However, there were no significant differences in the levels of total nitrogen (TN) and total phosphorus (TP) between the urban and peri-urban watersheds. The results showed that the urban watershed had higher terrestrial humic-like C1 (39%) and lower protein-like C3 (30%) than the peri-urban watershed, while the peri-urban watershed showed an inverse trend (33% and 37%, respectively). The results also revealed that the DOM fluorescent indices were significantly different between the peri-urban and the urban watersheds. Redundancy analysis (RDA) was applied to evaluate the correlation between DOM fluorescent indices and water quality parameters (i.e., COD, TN, TP, DOC). It revealed that the pollution sources and water quality correlated to the fluorescent indices and the C1-C3 DOM fluorescent components. A significant linear relationship between COD and C2 was found in both watersheds, suggesting that C2 might be a good COD indicator. Our results suggest that the distinctive DOM composition between the watersheds could be attributed to different human activities at both sites. The correlation between DOM fluorescent components and water quality can be assessed by the EEM-PARAFAC method, indicating considerable potential for the use of this technique to monitor surface water quality.