Atomic force microscopy-infrared (AFM-IR) spectroscopy is a powerful new technique that can be applied to study molecular composition of cells and tissues at the nanoscale. AFM-IR maps are acquired using a single wavenumber value: they show either the absorbance plotted against a single wavenumber value or a ratio of two absorbance values. Here, we implement multivariate image analysis to generate multivariate AFM-IR maps and use this approach to resolve subcellular structural information in red blood cells infected with Plasmodium falciparum at different stages of development. This was achieved by converting the discrete spectral points into a multispectral line spectrum prior to multivariate image reconstruction. The approach was used to generate compositional maps of subcellular structures in the parasites, including the food vacuole, lipid inclusions, and the nucleus, on the basis of the intensity of hemozoin, hemoglobin, lipid, and DNA IR marker bands, respectively. Confocal Raman spectroscopy was used to validate the presence of hemozoin in the regions identified by the AFM-IR technique. The high spatial resolution of AFM-IR combined with hyperspectral modeling enables the direct detection of subcellular components, without the need for cell sectioning or immunological/biochemical staining. Multispectral-AFM-IR thus has the capacity to probe the phenotype of the malaria parasite during its intraerythrocytic development. This enables novel approaches to studying the mode of action of antimalarial drugs and the phenotypes of drug-resistant parasites, thus contributing to the development of diagnostic and control measures.