Polycyclic aromatic hydrocarbons (PAHs) are listed as the priority pollutants. It is difficult to resolve effectively the peaks of PAHs by conventional one-dimensional fluorescence spectroscopy due to its low content and the overlapping fluorescence peaks. In this paper, the two-dimensional (2D) fluorescence correlation spectroscopy will be applied to the analysis of highly overlapped fluorescence spectra of the mixed solution of anthracene, phenanthrene and pyrene. According to the research goals, three mixed systems and a total of 27 samples, are to be prepared with different concentrations of three PAHs. Concentrations of three PAHS are monotonically increasing or decreasing in each mixed system. Then the 2D fluorescence correlation spectrum of each mixed systems will be calculated under the perturbation of the concentration of anthracene, phenanthrene and pyrene in solution. There are seven strong autopeaks at 425, 402, 381, 373, 365, 393 and 347 nm in synchronous 2D correlation spectrum. The fluorescence peak of phenanthrene at 347 nm is uncovered in three mixed systems, so the band at 347 nm is to be used as clues for further assignment. According to positive or negative cross peaks at 347 nm in synchronous 2D correlation spectrum, we can know that the peaks at 402, 381, 425 and 452 nm are assigned to anthracene, the peaks at 373 and 393 nm are assigned to pyrene, and the peaks at 365, 356 and 347 nm are assigned to phenanthrene. The fluorescence peak of phenanthrene at 385 nm is shown in asynchronous 2D correlation spectrum; it means the spectral resolution of asynchronous spectrum is better than the synchronous spectrum. The results are that it is feasible to analyze serious overlapping multi-component PAHs using two-dimensional fluorescence correlation spectroscopy, which can be extended to the detection of other pollutants in the air.