This study proposes a new technique that combines three-dimensional fluorescence spectra with algorithm combination methodology (ACM) to address issues associated with complex components of oil pollutants and the difficulty in identifying their overlapping spectra. By combining alternating trilinear decomposition (ATLD), self-weighted alternating trilinear decomposition (SWATLD), and parallel factor analysis (PARAFAC), ACM realizes the complementary advantages of using three algorithms. First, using carbon tetrachloride as the target contaminant, a three-component mixed solution of diesel, gasoline, and kerosene with different concentrations are prepared. Then, the three-dimensional fluorescence spectra of the mixed solution are measured using a F-7000 fluorescence spectrometer. Blank deduction and missing data recovery-principal component analysis are then employed as pretreatment methods to eliminate the scattering interference. Finally, ACM is used to decompose the three-dimensional spectral data matrix. Results are compared with the three separate algorithms for component analysis, revealing that ACM is insensitive to component concentration. The average recoveries for diesel, gasoline, and kerosene are 96.68%, 97.83% and 97.11%, respectively, which further indicated that this method is more universal and can be used for the qualitative and quantitative analyses of contaminants in oil mixtures.