AbstractThree multivariate calibration methods, including classical least square with nonzero intercept (CLS), principal component regression (PCR) and partial least square (PLS), have been used for the determination of pyritinol dihydrochloride in the presence of its degradation product. The CLS, PCR and PLS techniques are useful in spectral analysis because the simultaneous inclusion of many spectral wavelengths instead of the single wavelength used in derivative spectrophotometry has greatly improved the precision and predictive abilities of these multivariate calibrations. A training set was constructed for the mixture and the best model was used for the prediction of the concentration of the selected drug. The proposed procedures were applied successfully in the determination of pyritinol dihydrochloride in laboratory‐prepared mixtures and in commercial preparations. Pyritinol dihydrochloride was analysed with mean accuracies 99.99 ± 0.905, 99.91 ± 0.966 and 99.92 ± 0.962 using the CLS, PCR and PLS methods respectively. The validity of the proposed methods was assessed using the standard addition technique. The proposed procedures were found to be rapid and simple and required no preliminary separation. They can therefore be used for the routine analysis of pyritinol dihydrochloride in quality‐control laboratories. Copyright © 2009 John Wiley & Sons, Ltd.