We have developed quantitative structure-activity relationship models for a series of hetero aromatic tetrahydro-1,4-oxazine derivatives exhibiting antioxidant as well as squalene synthase inhibitory activities using genetic function approximation and genetic partial least squares regression techniques. The independent variables used for the model development include descriptors belonging to various categories, viz. molecular shape analysis descriptors, Jurs spatial descriptors, shadow indices, electronic parameters and quantum chemical descriptors (Mulliken charges of the common atoms shared by the 22 oxazine derivatives). The developed models were validated internally by leave-one-out and leave-many (1/7th)-out cross-validation techniques. The two pairs of models developed for the two different responses, besides being robust (as observed from the results of randomization) are quite efficient in terms of their predictive ability (as observed from their cross-validated Q(2) and r(m(LOO))(2)). The developed models suggest that the antioxidant activity of the oxazine derivatives is primarily controlled by the electrophilic nature of the molecules together with the charges on the phenolic hydrogen and the steric volume occupied by the molecules. Again the models for squalene synthase inhibitory activity suggest that the charges on the hetero aromatic nucleus as well as the charged surface area of the molecules and their size govern this response.