The search for alternative environmentally friendly refrigerants have never been so crucial with the increasing demand for effective cooling of increasing miniaturization of our heat exchanging devices in the ever expanding air-conditioning and refrigeration industry. Although propane (R290) and ammonia (R717), natural refrigerants, have been around for decades, their two-phase thermal performance in small channels has yet to be fully investigated. Predictions of the heat transfer using correlations developed based on past experimental data have shown poor agreements, with more correlations being developed to date. This research was done to investigate the optimized conditions for the two-phase boiling heat transfer coefficient of R290 and R717 where the contributions from nucleate boiling and forced convective are represented explicitly. Multi-objective Genetic Algorithm (MOGA) is utilized for the simultaneous maximization of nucleate boiling and forced convective, two conflicting phenomena – the former generally significant in the low vapor quality region while the latter in the high quality region. A superposition correlation is used as it sums up both contributions. Two phased-out refrigerants, R134a and R22 are also being research here for comparison purposes. The range of MOGA design parameters set for mass flux, G, is between 100 - 300 kg/m2.s, heat flux q between 5 - 30 kW/m2 and vapor quality, x for 0.0009 - 0.9. The optimization is done for 3 mm channel diameter with saturation temperature at 10ËšC. The optimized results showed a strong contribution of each nucleate boiling and forced convective for R717 with increasing vapor quality, compared to the other three refrigerants. The optimized value of the total heat transfer coefficient for R717 could reach up to 90 kW/m2.K and for R290 up to 12 kW/m2.K compared to R134a and R22 at 6 kW/m2.K and 5 kW/m2.K respectively. At lower vapor quality, the nucleate boiling contributes more to the total heat transfer coefficient, and suppressed due to forced convective as the vapor quality reaches middle range. The theoretical results indicate the potential of R717 and R290 as replacement refrigerants for R22 and R134a with further verifications to be done with correlations not using the superposition method.