The performances of an Alternating Least Squares (ALS) multivariate curve resolution method, recently developed, and SQUAD - a traditional least-squares curve fitting method - are studied for different sets of simulated and real data of the acid-base equilibria of polynucleotide polycytidylic acid. The influence of the lack of selectivity, of the polyelectrolyte effect and drift baseline on the recovered pH concentration profiles and pure species spectra, is investigated. The SQUAD method gives better results when there are no polyelectrolite effects and the systems completely lack any selectivity. Converseley, for systems including high polyelectrolyte effects and some selectivity, the ALS method performs better than the SQUAD method.