Abstract: Three-dimensional excitation-emission matrix spectroscopy (EEMs), parallel factor (PARAFAC) analysis, and hierarchical clustering analysis (HCA) with classification and regression tree (CART) were applied for spectroscopic data to characterize fluorescent components of dissolved organic matter (DOM) in the surface sediment samples, which were collected in rural, urban and city sections of the Baitapu River, a typical urbanization river in Northeast China. The spatial characteristics of fluorescence components were analyzed, the key factors of DOM characteristics identified, and the pollution sources distinguished. The results showed that six DOM fluorescent components were extracted from DOM of Baitapu River sediment, which contained fulvic acid (C1), microbial metabolites (C2), humic acid (C3), tryptophan-like (C4), tyrosine-like (C5) and phenolic (C6) materials. The fluorescence concentration of DOM was the highest in the town section of the Baitapu River, followed by the city section and the lowest in the rural section. The DOM fractions in the rural section was dominated by autochthonous origins, while the DOM fractions in the town and urban sections by terrigenous and autochthonous substances. The pollution degree of bottom sediment of the Baitabu River was ordered as town section>urban section> rural section, and the bottom sediment DOM showed the differentiation characteristics of rural section, town section and urban section. C1 and C5 were two key factors to recognize the DOM characteristics of sediment.