Classification of data originating from complex process and biological systems is challenging owing to the presence of multivariate and highly nonlinear interactions between variables. Patterns, difficult to distinguish using decision boundaries or available discriminating rules, can be separated based on unique inter-relations among the feature vectors. Given the complex nature of such systems, the variable interaction models are difficult to establish. Genetic programming (GP), a data-driven evolutionary modeling approach, is suggested here to be a potential tool for designing variable dependency models and exploiting them further for class discriminant analysis. Thus, this paper proposes a new GP model based classification approach. The approach is applied on illustrative data sets, and its performance is benchmarked against well-established linear and nonlinear classifiers such as LDA, kNN, CART, ANN, and SVM. It is demonstrated that GP based models can play an effective role in classification of data into multiple classes.