A critical comparison of the selected derivative principal component analysis (PCA) methods on the absorbance matrix data concerning the complexation equilibria between o-CAPAZOXS and Cd(2+), Pb(2+) and Zn(2+) or m-CAPAZOXS and Cd(2+), Co(2+), Ni(2+) and Zn(2+) or p-CAPAZOXS and Cd(2+) and Zn(2+) at 25 degrees C is provided. As the number of complex species in a complex-forming equilibria mixture is an important step in spectral data treatment, the nine selected index functions for the prediction of the number of light-absorbing species that contribute to a set of spectra is critically tested by the PCA. An improved identification with the second SD(AE) or third derivative TD(AE) and derivative ratio function ROD(AE) for the average error criterion AE is preferred. After the number of various complexes formed the stability constants of species ML, ML(2) (and ML(3), respectively) type log beta(11), log beta(12) (and log beta(13), respectively) for the system of o-CAPAZOXS (ligand L) with the metals (the standard deviation s(log beta(pq)) of the last valid digits is given in brackets) Cd(2+) (6.39(5) and 11.51(9)), Pb(2+) (4.24(2) and 9.01(2)) and Zn(2+) (5.18(7) and 9.06(10)) and for the system of m-CAPAZOXS with Cd(2+) (6.59(20) and 11.51(32)), Co(2+) (7.19(6) and 12.19(8)), Ni(2+) (7.64(7) and 13.39(12)) and Zn(2+) (4.83(3) and 9.57(3)) and for the system of p-CAPAZOXS with Cd(2+) (6.44(5), 10.99(10) and 14.57(25)) and Zn(2+) (6.84(16), 13.05(29) and 18.74(43)) at 25 degrees C are estimated using SQUAD(84) nonlinear regression of the mole-ratio spectrophotometric data. The computational strategy is presented with goodness-of-fit tests and various regression diagnostics capable of proving the reliability of the chemical model proposed.