The fluorometric determination of the ground-state dissociation constant k(d) of a complex between ligand and titrant with 1:n stoichiometry in the presence and absence of excited-state association is discussed. This report extends the results of a previous study (J. Phys. Chem. 1994, 98, 8585) where a 1:1 complex between ligand and titrant was assumed. For that case, it was demonstrated that adequate experimental design, i.e., by monitoring the fluorescence at the isoemissive point, could eliminate the interference of the excited-state association. In this report we use the direct parametric fit of the fluorometric titration to recover reliable estimates for K-d It is shown that in the presence of the excited-state association this approach yields the unique value of K-d if a priori knowledge of the excited-state rate constants is available. This method also offers an effective way for discriminating between competitive models (presence and absence of the excited-state association reaction with non 1:1 stoichiometry). The developed algorithms for the direct parametric fit of fluorometric titrations were explored by simulations and were applied to experimental fluorometric titration curves of new Mg2+ indicators Thio-H and terThio-H. The importance of the proper choice of experimental conditions-i.e., signal-to-noise ratio, excitation, and emission wavelengths-for the accuracy of the analysis and the quality of discrimination between competitive models is demonstrated.