The aim of the study was to evaluate using FTIR-spectroscopy the changes in bleached cotton, flax and kraft pulp papers after aqueous light bleaching and accelerated light ageing and to reveal influence some absorbance bands on paper brightness.

To solve this task the principal component regression (PCR) was applied to two sample sets. First set included only papers from cotton fibers and second one consisted of samples produced from different fibers. The FTIR spectra of the samples in the range 1460-–1840 cm^–1 were used as predictors.

Principal component analysis allowed to decompose the spectra and to get some interesting information about absorbance bands in the frequency range of interest. Thus for the paper samples from cotton fibers it was found out that intensities of the bands near 1512, 1540, 1577, 1590, 1656, 1690 and 1740 cm^–1 are quite different for treated and untreated samples. For the cotton, flax and kraft pulp papers the treatments resulted in changing of the same spectral bands with addition of the band near 1480 cm^–1. Some correlations in behavior of the bands values and changes in content of functional groups were also revealed.

Then the PCR was used to predict the paper brightness. Good models, with correlation between predicted and measured values about 0.95, were obtained for the both sample sets. It is worthy to notice that if in the first sample set only bands from functional groups vibrations gave main input to the paper brightness, for the second one a band connected with vibration of cellulose macromolecules situated in ordered ranges also had a strong influence.

Applying PCR to FTIR-spectroscopic data allowed to clear changing spectral curves in the range of C=O groups vibration at bleaching and ageing and to calibrate the models for prediction of paper brightness. Such approach has a real perspective for its application in paper restoration studies.