In this study, oil extracted from coconut fruits was optimized using Response Surface Methodology (RSM). Effects of roasting temperature and roasting time on the yield and oil quality were investigated. Thirteen experimental runs applying a central composite design combined with RSM was employed. The parameters measured were oil yield, free fatty acid, colour, refractive index, specific gravity and pH. Statistical analysis with response surface regression showed that the oil yield, free fatty acid, colour, refractive index, specific gravity and pH of coconut oil were significantly (p<0.05) affected with roasting temperature and time. Based on response surface, optimum conditions were roasting temperature of 110 oC and roasting time of 17 min. Analysis of variance indicating that the model was adequate for representing the experimental data. The treatments resulted in oil yield ranging from (31.28 to 42.37%), free fatty acid (1.86 to 5.34%), colour (0.46 to 0.66 abs), refractive index (1.47 to 1.51), specific gravity (0.98 to 1.04) and pH (6.28 to 6.94). Oil extracted from coconut was successfully optimized using RSM. The regression models obtained has provided a basis for selecting optimum process variables for the recovery of oil using solvent extraction.