Using of the supercritical fluid extraction to extract natural products is one of the methods that has played an important role in the development of the industry. Advantages of SFE are: law solvent consumption, high security, reasonable cost, law and safe wast, suitable for sensitive materials of heat, lack of water and non-corrosive, acceptable extraction time. Summer Savory belongs to lamiace and Thymol is one of its main components. In this study, extraction of Thymol from Summer Savory was performed by using supercritical carbon dioxid. Soxhlet method with ethanol was considered as the base case of 100% recovery of Thymol. Extracted essential was analyzed by High –performance liquid chromatography (HPLC). Supercritical carbon dioxid extreaction was performed by ethanol as co-solvent and particle size with mesh size 20-35 and 30 min as static time. The Design of experiment was carried out with response surface method (RSM) and using Minitab software, and considering four independent parameters temperature (35-75°C), pressure (10-30 MPa), dynamic time (40-120 min) and flow rate (0.8-2.4 ml/min). The response surface analysis showed modeling of recovery with a second order polynomial equation, the linear and quadratics term of temperature, pressure, dynamic time, flow rate and the interaction between temperature- flow rate, dynamic time-flow rate had significant effect on the proposed model of Thymol recovery. R2 and modified R2 of the model are 97.15% and 94.65% respectively. It shows laboratory data was fitted well with prposed model. Optimal values of extraction parameters in this experiment were, temperature 42.67°C, pressure 21.31 MPa, dynamic time 109.49 min and flow rate 1.8 ml/min, that under these conditions, recovery 49.95% was achieved. At the end, the supercritical carbon dioxid extraction data was modeled by artificial neural network method and the result were compared with result of RSM. Its determination coefficient (R2 ) was equal 99.59%. Optimal number of hidden layer neurons was 6 and the BR training algorithm was known as the best performance of the algorithms. With comparing the result of R2 in two methods (RSM ANN) was found ANN performed better than RSM. Keywords Supercritical carbon dioxid extraction, Thymol, Summer Savory, Ethanol, Respons Surface Methodology, Artificial Neural Network