Diesel/natural gas dual-fuel engines were studied numerically using Latin hypercube sampling (LHS) and multi-objective Pareto optimization. The combustion characteristics were analyzed and performance metrics studied as a function of the exhaust gas recirculation (EGR) fraction, compression ratio, injection timing, and natural gas fraction (NGF). The fundamental effects of the NGF on the in-cylinder pressure, peak temperature, brake torque, and brake specific fuel consumption (BSFC) were investigated. Regression coefficients were obtained using LHS to describe the sensitivity of the brake torque, BSFC, and brake specific NOX (BSNOX) emission levels to the operating parameters. A response surface was found using Pareto optimization. The dominant parameters affecting the brake torque, BSFC, and BSNOX were the compression ratio and EGR fraction. (C) 2014 Elsevier Ltd. All rights reserved.