Recently several triazine derivatives were identified as excellent cytotoxic agents against non-small cell lung carcinoma by our group. QSAR model for prediction of biological activity of triazine derivatives against non-small cell lung carcinoma cell line (A549) is needed to construct selective inhibitors for lung cancer. Twenty four models were constructed using genetic function approximation algorithm. The best of these models was chosen based on its statistical validation parameters where the R2 value was found to be acceptable (0.98).The developed model was based on four molecular descriptors; two fast descriptors and two VAMP electrostatics descriptors. External validation of the model was governed by calculating the residual values for test set. Further external validation is investigated by calculating the biological activity of four new triazine derivatives synthesized by our group in a previous contribution from our laboratory. Our developed model is proved to have high predictive and diagnostic abilities and can distinguish different stereoisomers. The accuracy of the 3D structures used affects the model quality.