An onsite technique for determining drug integrity in sub-Saharan Africa is needed to ensure drug integrity and enhance public health. This current study presents the application of handheld NIR spectroscopic and multivariate techniques for the accurate identification of unexpired drugs from expired ones. A total of 150 drugs comprising 75 drug samples each of antimalarial (40 unexpired and 35 expired) and antibiotics (40 unexpired and 35 expired) were used in the study. Principal component (PC) analysis was used to extract relevant information from the spectral fingerprint and pre-processed using different techniques comparatively to observe the best cluster trends. The performance of three multivariate algorithms: RF, SVM, and PLS-DA were compared after optimization by cross-validation. The results revealed that SVM and PLS-DA were superior with an identification rate for both antimalarial and antibiotic authenticity prediction above 98% at 5 PCs in both the prediction set and calibration set. For simultaneous prediction of expired and unexpired drugs, we achieved a 100% identification rate. Generally, the results show that handheld NIR spectrometers coupled with smartphone devices could successfully be used to identify unexpired antimalarial and antibiotic drugs from expired antimalarial and antibiotic drugs for effective quality assurance in poor-resource countries. This offers positive feasibility for an affordable and user-friendly approach to reducing drug fraud in Africa.