Ratio estimators have been used in survey sampling for estimating population mean when the population mean of an auxiliary variable is known. However, as the auxiliary information is not always available we cannot gain any benefit from using ratio estimators to increase the efficiency of the population mean estimator. New ratio estimators for population mean obtained using the quartile function of an auxiliary variable using double sampling have been proposed based upon the modified ratio estimators proposed by Subramani and Kumarapandiyan (2012). We propose the use of double sampling in order to estimate the unknown quartile function of an auxiliary variable. A simulation study has been conducted to compare these new ratio estimators with the classical ratio estimator using a percent relative efficiency using double sampling. The results show that the proposed estimators perform better than the classical ratio estimator using double sampling.