The spread of antimalarial drug resistance and lack of effective vaccines into consideration drive significant role to develop novel antimalarial agents for the treatment of malaria. ThePlasmodium falciparum deoxyuridine 5′-triphosphate nucleotidohydrolase (dUTPase) has been identified as a promising target for new drug development. In this study, we report a receptor-dependent mixed 2D- and 3D-QSAR model using the LQTA-QSAR methodology on a series of dUTPase inhibitors. QSAR models were constructed by a training set of 55 compounds with specific stereochemistry and an external set of 18. The model satisfied a set of rigorous validation criteria and performed well in the prediction of an external test set. The proposed model also checked for free from chance correlation, reliability, and robustness by y-randomization and leave-N-out tests. Final model provided the following statistics (R 2 = 0.85,Q LOO 2 = 0.78 and Q ext 2 = 0.76) with only 12 descriptors and 3 latent variables. We showed that a combination of 2D classic and 3D molecular field descriptors could lead to relevant QSAR models, that in comparison with model solely generated by 3D descriptors has improved and illustrated more robustness and prediction ability. Visualization of the descriptors of the best model helps us to interpret the model from the chemical point of view, supporting the applicability of this new approach in rational drug design.