In this study, experimental supercritical CO2 extraction of beta-carotene from leaf skin of Aloe vera (Aloe barbadensis miller) was investigated and modeling of this system was developed in order to predict the extraction yield as a function of effective operating variables. The results of beta-carotene extraction via supercritical CO2 were compared with the conventional Soxhlet extraction from almost 5g of Aloe vera skin with 100 ml petroleum ether as solvent. The shrinking core model was applied in the simulation which is similar to irreversible desorption of a solute from a porous adsorbent. The modeling predictions are very well compatible with the experimental data. The beta-carotene extraction yield was determined as a function of temperature, pressure, CO2 flow rate and dynamic extraction time. The genetic algorithm (GA) was used to determine the optimal operating conditions in which the maximum amount of beta-carotene extraction yield was obtained. (C) 2012 Elsevier B.V. All rights reserved.