In this paper, a new estimate of exponential type of auxiliary information to help simple random sampling without replacement of the finite population mean is introduced. This new estimator with a few other estimates using two real data sets are compared with the mean square error.