This paper proposes a statistics-based process control method for small-batch production, specifically developed to address the insufficient sensitivity of traditional Shewhart control charts in handling small-sample data scenarios. By integrating multidimensional statistical methods including F-tests, T-tests, and process shift analysis, a predefined-limit control chart model tailored for small sample sizes was developed. Based on the first-pass yield (FPY) of the process and historical data, this model reverse-engineers the sample standard deviation and utilizes the “3-sigma” principle to establish the control limits, thereby reducing reliance on extensive subgroup data. The research results demonstrate that this method reduces the minimum required number of subgroups from 20 to 5, significantly improving efficiency, while maintaining high predictive accuracy. It enables real-time monitoring of production processes and evaluation of quality metrics. This approach provides a highly efficient and reliable solution for small-batch production, facilitating timely detection of deviations and ensuring product quality meets specified requirements.