In this work, we employed the wavelength selection method to construct Partial Least Squares (PLS) models for simultaneous quantification of Amlodipine (AML) and Indapamide (IND) in pharmaceuticals. Three wavelength selection methods, including sensitivity ratio (SR), backward variable elimination (BVE), and genetic algorithm (GA), were utilized. The results demonstrated that the SR method selected the highest number of wavelengths, while the GA method selected the fewest. We also developed calibration models including PLS, SR-PLS, BVE-PLS, and GA-PLS. The statistical values for both the calibration and validation datasets were favorable. Evaluation based on recovery rate indicated that the recovery results were within the range of 90-110% for both AML and IND. The repeatability also yielded satisfactory outcomes, with RSD values <
RSD calculated using the Horwitz function. The quantification results of AML and IND fell within the range of 90-110% compared to the labeled content, adhering to the regulations of the Vietnamese Pharmacopoeia V.