One of the most serious problems that can occur in classical quantitative analysis is the presence of one or more spectral interferent-chemical species which affect the instrument response and which are unaccounted for in the calibration process. Advances in chemometric methods allow quantitative analysis in the presence of unidentified interferents if a threeway experimental data matrix is available for each sample. This property is the so-called "second-order advantage" 1 and it is based on the earlier work done in the psychometrics field. 2,3 The use of this advantage in analytical chemistry was proposed by Ho et al. 4 (rank annihilation factor analysis, RAFA), for the multicomponent analysis of fluorescent mixtures using excitation-emission matrix (EEM). Later, Lorber 5 and Sanchez and Kowalski 6 proposed new and simple solutions for the method of Ho et al. and now it is called the generalized rank annihilation method (GRAM).At present, analytical chemistry laboratories have instrumentation that easily generates multidimensional data structures of experimental data for each sample. Fluorescence is a particularly interesting technique because it allows, in a very straightforward manner, the acquisition of this type of information, of which the most typical example is the excitation-emission matrix (EEM). Several algorithms that use second-order data are described in the literature. [7][8][9] Of these, the generalized rank annihilation method (GRAM) 6,11,12 has interesting properties since it only requires two data matrices for quantification: one from the calibration sample and the other from the unknown sample. Recently, it has been proved that GRAM predicts results that are similar to those from the parallel factor analysis (PARAFAC) when only two samples are used, 9 so both methods can be useful as calibration methods for second-order data. But, unlike PARAFAC, GRAM is a non-iterative method which tries to implement the second-order advantage. Second-order calibration using GRAM has several distinct advantages. 13The most pronounced advantage is that it makes possible the quantification of the analyte(s) of interest in samples containing interferents. Those interferents can be completely unknown, that is, it is not necessary that they have been incorporated in the standards used for calibration. Another advantage is that it is possible to recover the individual instrument responses of the analyte(s) of interest, e.g., the excitation and emission spectra can be obtained for the individual analyte(s). This makes it possible to perform qualitative analysis and to check the identity of the analytes. The Kowalski group has proposed the trilinear decomposition (TLD) method. 14,15 In this work we focused our attention on the ability of GRAM for analyzing the second-order data.Fluorescence spectroscopy is a versatile tool mainly used because of its selectivity and sensitivity. Fluorescence spectra can be recorded in difference modes such as emission, excitation, synchronous and excitation-emission matrix (EEM). T...