A rapid, non-destructive and reagent-free infrared spectroscopy combined with Partial Least Square (PLS) has been developed for the dexamethasone quantification in joint-pain killer traditional herbal medicine (THM). The main aim of this study is to select the best wavenumbers that are capable of providing the high coefficient of determination (R2), low values of Root Mean Square Error of Calibration (RMSEC), Root Mean Square Error of Cross Validation (RMSECV) and predictive residual error sum of squares (PRESS). Finally, wavenumbers 3646, 3642, 2461, 2453, 2432, 2406, 2229, 2209, 2197, 2097, 2092, 2064, 2059, 2047, 2026, 2009, 1969, and 1513 cm-1 were selected for the prediction of dexamethasone in the joint-pain killer traditional herbal medicine. The correlation between the actual values of dexamethasone determined in joint-pain killer traditional herbal medicine using infrared spectroscopy combined with PLS revealed the R2 values of 0.9988. The RMSEC values obtained 0,009455. The PRESS and RMSECV value obtained as the results of cross-validation model selection for dexamethasone in joint-pain killer traditional herbal medicine were 0,0022721.00 and 0,02902, respectively. The high value of R2 and low value of RMSEC, RMSECV and PRESS indicated that this method had high accuracy and precision in a validated condition for the dexamethasone quantification in the joint-pain killer traditional herbal medicine. These results indicated that infrared spectroscopy combined with PLS can be an alternative method for the dexamethasone determination in joint-pain killer traditional herbal medicine.his part contains English version of the The abstract presents background, method of the research/ literary study and discussion. The abstract consist of maximum 300 words. All sentence must represent the core of research presented in good structure of sentences.