The extraction efficiency of naphthalene from an organic phase depends essentially on the choice of extractant. It can be enhanced through the optimization of operating parameters. The organic phase studied was a fuel oil rejected at the process end of the coke gas treatment in an industrial plant of Annaba (Algeria). Themixture ofmethanol.phenol has been selected as extractant for its best efficiency. The amount of naphthalene in the oil phase was analyzed by refractometry and verified by infrared spectroscopy. The temperature, the density of methanol.phenol mixture and the quantity of extractant, were considered as the main parameters which affect the performance of extraction. Response surface methodology was applied to investigate the yield of extraction. A Box-Behnken design was used to produce models for objective responses based on parameters that have significant probabilities. It was developed by using these three main parameters at three levels. Analysis of variance showed that temperature and the volume ratio are the most affecting parameters on extraction yield. The developed regression equations described interdependencies between the input parameters and yield of extraction as responses. The results that were treated statistically were very interesting and allowed to deduce the optimum conditions for the best extraction efficiency which was in order to 80.38%. The application of extraction in these conditions has permitted to regenerate the fuel oil rejected by the industrial unit for other uses. The phenol.methanol mixture containing naphthalene can also be regenerated (91.8% of methanol and 94.2% of phenol) and more than 1 ton of pure naphthalene per day can be recovered. This could be very important for the industrial unit on decreasing the cost of the fuel consumption and more profits by producing pure naphthalene.