Control chart is one of the most commonly used statistical tools for quality control and improvement. If the process mean and standard deviation are not given or unknown, most Shewhart control charts require sufficient sample data before the control chart can be established. However, in certain industries or processes, it may not be practical to collect adequate amount of data at the beginning of the manufacturing process to build the trial control chart in Phase I. For quality improvement in such or similar processes, some authors developed self-starting control charts for short-run production, e.g. t chart, Q chart, EWMA t chart/Q chart, CUSUM t chart/Q chart. This thesis studies the performance of some short run control charts for monitoring process variances. Numerical simulations are using in this study. The results of the numerical experiments are extensively tested for different combinations of process lengths and starting points of process shifts.