A meta-model based multi-objective optimization endeavor was undertaken in the present work to investigate the potential of the off-line model based calibration technique to extend the actual CRDI-CNG dual-fuel experimental investigations in determining the possibility of unearthing viable potential trade-off domains hitherto unexplored by the constraints of resources, cost and time warranted by an experimental investigation. For the ensuing optimization study, CNG energy share, fuel injection pressure and load have been used as the decision variables while PM, NHC and BSFC were chosen as the output variables to be optimized. In absence of a closed form correlation between the participating variables under study, the explicit characterization capability of the Gene Expression Programming technique was harnessed. The appropriate GEP based meta-models were adopted from a previous study correlating the identical system output responses for the same set of decision variables of interest in the present study. Genetic algorithm was chosen as the optimization routine in the present study in view of its promising potential of extremely fast convergent speed, diversity of optimal solutions and simplicity of operation. Experimental validation of the obtained solutions pertaining to the desired objectives were carried out by actual experimentation. The present optimization endeavor was able to better the best vantage in category of the desired objective of minimum fuel consumption and exhaust emissions, obtained not only as compared to baseline diesel operation comprehensively but also was superior than the actual CRDI-CNG strategy during actual dual-fuel operation corresponding to actual experimentation.