A quantitative structure–activity relationship (QSAR) analysis has been performed on a data set of 194 artemisinin analogs for antimalarial activity. Several types of descriptors including topological, spatial, thermodynamics, information content, lead likeness, and E-state indices have been used to derive a quantitative relationship between antimalarial activity and structural properties. A systematic approach of zero tests, missing value test, simple correlation test, multicollinearity test, and genetic algorithm method of variable selection was used to generate the model. Statistically significant model (r2 = 0.845, q = 0.799, F-test = 53.40) was obtained with the descriptors like molecular connectivity indexes, E-state index, length-to-breadth ratio of compounds, MLog P, HOMO, electron density, Balabans topological index, and strain energy of the molecules. The robustness of the QSAR models was characterized by the values of the internal leave one out cross-validated regression coefficient (q) for the training set and determination coefficient in prediction, q for the test set. The value of q = 0.876 for the test set; revealed good external predictability of the QSAR model. Also, for an external data set (validation set) of four artemisinin analogs, the QSAR model was able to predict the antimalarial activity very well in comparison to experimental values. The model was also tested successfully for external validation criteria. The QSAR model developed in this study should aid further design of novel potent artemisinin derivatives. Copyright © 2009 John Wiley & Sons, Ltd.