Correlations for bubble point pressure, solution gas-oil ratio (GOR), oil formation volume factor (OFVF) (for both saturated and undersaturated crude) and viscosity (for both saturated and undersaturated crude) have been developed for Indian (west coast) crude using Artificial Neural Networks (ANN). Detailed comparison has also been made with various important correlations currently available in the literature. Sensitivity analysis of the developed models was also performed to determine the relative importance of various input parameters. The training scheme used here is different from those used previously for developing ANN models. Bayesian regularization technique was used to ensure generalization and prevent over fitting. Also genetic algorithm (real coded with parent-centric crossover) was used coupled with a local optimizer (Marquardt-Levenberg) to obtain the global optimum network weights. It was found that the developed models outperformed most other existing correlations by giving significantly lower values of average absolute relative error for the parameters studied. This study shows highly favorable results which can be integrated in most reservoir modeling software. © 2010 Elsevier B.V.