With the increasing complexity of analytical data, traditional data processing methods for spectral analysis can not meet the analysis needs. The chemometrics uses methods of mathematics, statistics and computer sciences to design optimum chemical measurement procedures and to provide maximum information about fluorescence spectrum by analyzing data. In the present paper, the main kinds of three-dimensional fluorescence spectrum, which are widely analyzed by chemometric methods, were illustrated at first. Secondly, an overview of the chemometric methods such as statistical methods, pattern recognition, neural network and the second-order calibration methods is presented. And the applications of these methods are followed by. Finally, the prospect of chemometric methods applied in multidimensional spectral analysis is discussed.