This paper discusses the use of weighted variance (WV) in setting up the limits of the exponentially weighted moving average (EWMA) chart for the monitoring of the mean of a process from a skewed population. This chart, called the WV-EWMA chart hereafter, reduces to the standard EWMA chart when the underlying distribution is symmetric. The Type-I and Type-II errors of the WV-EWMA chart are compared with that of the existing charts for skewed populations. Simulation results show that the new method gives a considerable improvement over the existing methods when the underlying distribution is skewed.