Malaria is an infectious disease that represents a major concern worldwide. This situation has created the necessity for the rapid development of drugs for its treatment. Cheminformatics has positioned itself as a helpful tool for accelerating drug design by conjugating chemistry with mathematical and computational methods. The present study proposes a quantitative structure-activity relationship (QSAR) model to assist the design of antimalarial drugs based on L-mannitol. The study was divided into three stages. During the first stage, the most suitable chemical descriptors for the model were identified from a database consisting of 19 molecules derived from L-mannitol and 25 descriptors from the categories molecular properties, constitutional indices, and functional group counts. Then, the data of the selected descriptors were analyzed through a multiple linear regression to derive the mathematical equation for the model. Finally, cross-validation was performed on the model to determine its predictive ability. The results showed that the ionization potential, the number of secondary sp3 carbons in the molecule, and the number of donor atoms for hydrogen bonds are crucial molecular features for explaining the inhibitory activity of L-mannitol-based compounds over the Plasmepsin II protease. Additionally, the values obtained for the correlation coefficient (R2 = 0.81) and cross-validated correlation coefficient (Q2 = 0.73) suggest a good predictive ability of the model.