In this study, organic matter distributions and concentrations at various stages of an urban wastewater treatment plant in Hefei, Eastern China, were analyzed from November 2022 to June 2023 using UV-Vis absorption and three-dimensional fluorescence spectroscopy. Six components identified via excitation-emission matrix and PARAFAC analysis revealed that components related to tryptophan (components 1, 3, and 6) had strong correlations with COD concentrations, with Pearson correlation coefficients of 0.656, 0.447, and 0.674, respectively. Analysis of fluorescence and UV-Vis absorption parameters indicated a reduction in organic matter content, increased humification, and a shift from exogenous to endogenous organic substances throughout the sewage treatment process. Notably, the humification index showed the highest correlation with COD levels (-0.834). On this basis, Utilizing Monte Carlo-Uninformative Variable Elimination-Partial Least Squares (MC-UVE-PLS) for characteristic wavelength extraction from normalized fluorescence and absorption spectra, a COD characteristic fusion spectral analysis model was developed. The results show that there is a good agreement between COD concentrations obtained based on feature fusion spectral analysis and COD true values obtained by the potassium dichromate method. The coefficient of determination between COD predicted values and the true values in the testing set reached 0.9725, and the root mean square error was only 10.51 mg/L. These findings suggest the efficacy of using UV-Vis absorption and three-dimensional fluorescence spectroscopy for direct COD tracking and detection in wastewater treatment processes without any pretreatment.