In automotive industry it has been a challenge to retain diesel-like thermal efficiency while maintaining low emissions. Numerous studies have shown significant progress in achieving low emissions through the introduction of common-rail injection systems, multiple injections and exhaust gas recirculation and by using a high octane number fuel, like gasoline, to achieve adequate premixing. On the other hand, low temperature combustion strategies, like HCCI and PCCI, have also shown promising results in terms of reducing both NOx and soot emissions simultaneously. With the increasing capacity of computers, multi-dimensional CFD engine modeling enables a reasonably good prediction of combustion characteristics and pollutant emissions, which is the motivation behind the present research. The current research effort presents an optimization study of light-duty compression ignition engine performance, while meeting the emission regulation targets. A numerical optimization study was carried out on a light-duty, single-cylinder, compression ignition engine, fueled with a PRF87 gasoline surrogate, at a full load operating condition. The simulations were performed using a Non-dominated Sorting Genetic Algorithm II (NSGAII) code coupled to a multi-dimensional CFD code, KIVA3V-Chemkin. The goal of this study was to reduce six objectives simultaneously, which are NOx, soot, carbon monoxide and unburned hydrocarbon emissions, indicated specific fuel consumption, and peak pressure rise rate. Six engine design parameters were chosen to vary while being optimized using the genetic algorithm code, which include the premixed fuel fraction, second and third injection amounts and timings, exhaust gas recirculation (EGR) ratio and boost pressure. The GA optimization results indicated that there are two distinct operating regimes depending on EGR ratio for the current engine operating conditions. The high EGR regime is characterized by late ignition timing using triple-injection operation, whereas, the comparatively low EGR regime was governed by two-stage combustion. The results showed that thermal efficiencies of more than 40% can be achieved with both strategies.