The process conditions that affect the percentage yield of oil from small castor seed was optimized for maximum extraction. In this study, surface response methodology (RSM) that employed a two-factor, five-level factorial central composite design (CCD) was used. Thirteen (13) experiments with different combinations of reaction time and reaction temperature ranging from 1hr (60mins) to 10hrs (600mins) and 60oC to 100oC respectively were also performed. A quadratic model that is polynomial in nature was obtained to predict the percentage oil yield for the small castor seed. Within the experimental variables range, the optimal conditions were found to be 8.68hrs (520.92mins) and 94.14oC respectively. These values were fitted into the quadratic polynomial model that gave rise to the optimum value of small castor seed oil yield to be 55.76% with a p-value less than 0.05. The coefficient of determination R2 was obtained as 0.9530 representing 95.30% of variation in the original data. Our result also gave an adjusted R2 value of 91.95% and predicted R2 value of 66.59% which indicate that the model explains 67% variation in predicting original observations. The result of this study showed clearly the percentage yield of oil that could be extracted from small seeded varieties of castor seed and the optimum yield value possible at a certain reaction time and temperature. The result also established the reliability of response surface methodology to model and optimize the expression of oil from small seeded varieties of castor seed.